Phoenix Made “No Section 8” Illegal in 2023. The Evidence Says That Helps — Slowly, and Only Where Someone Enforces It.
- A housing voucher is not a housing unit. In HUD’s own five-metro paired-testing study, landlord refusal ranged from 15% in Washington, DC to 78% in Fort Worth, Texas.1
- Refusal was worst in the best neighborhoods. In every city but DC, denial was more common in low-poverty areas — 85% of landlords tested in low-poverty Fort Worth neighborhoods and 82% in low-poverty Philadelphia neighborhoods turned vouchers away.1
- This is why vouchers expire unused. The share of voucher holders who manage to sign a lease before their search clock runs out fell from about 69% in 2000 to about 60% by 2019.2
- Source-of-income (SOI) laws do move the number. Voucher utilization rose 7 to 12 percentage points in jurisdictions that adopted them, relative to neighbors that didn’t.3
- But the gains arrive late. Effects on where voucher holders actually live took about three years to reach statistical significance in one study and about five in another.4
- And the effects are modest, not transformative. A study of laws with “stronger” enforcement language found no evidence they outperformed weaker ones — which points at enforcement in practice, not statutory text.4
- None of this is a randomized trial. Every causal estimate here is quasi-experimental — difference-in-differences and event studies on administrative data. That is the best evidence that exists on this question, and it is not the same thing as an RCT.5
- Arizona has no statewide protection. Phoenix (effective April 2023) and Tucson have city ordinances. Cross the city line and refusing a voucher is legal again.6
The Housing Choice Voucher program serves about 2.3 million households through more than 2,400 housing authorities, and it is the closest thing the United States has to a working answer to homelessness for people who can’t afford market rent.5 The waiting lists run years. And then, at the end of the wait, a piece of paper arrives that a landlord is — in most of the country — entirely free to throw away.
How often it happens, measured properly
The definitive measurement is a 2018 HUD-sponsored study by Mary Cunningham and colleagues at the Urban Institute. It is worth being specific about the method, because the method is why the numbers are trustworthy. Researchers screened roughly 341,000 online rental advertisements to isolate 8,735 units that a voucher could actually pay for under local payment standards, contacted about 3,780 landlords, and ran more than 4,000 phone and in-person tests across five metro areas. Where a landlord said they took vouchers, matched pairs of testers with equivalent profiles went further into the process to see whether treatment stayed equal.1
Fifteen percent of landlords in Washington, DC refused outright. Seventy-eight percent in Fort Worth did. In DC and Philadelphia, another 15% and 10% respectively said they weren’t sure of their voucher policy or would accept “under certain conditions” — answers the researchers flag as possible indirect discrimination.1
The geographic pattern is the part that should change how people read voucher data. Refusal was concentrated in low-poverty neighborhoods: 85% in low-poverty Fort Worth tracts, 82% in low-poverty Philadelphia tracts. When a dataset shows voucher households clustered in high-poverty areas, the intuitive reading is that they chose to stay near what they know. The testing evidence says something plainer happened first: in the neighborhoods with better schools, the phone call ended at the word “voucher.”1
Refusal also came early — on the first call, before an application, before a credit check, before anything a landlord could later characterize as a screening decision. And where landlords did say yes, some simply never showed up to the scheduled viewing. The New York City advocates who track this call it ghosting, and they report it has become more common as explicit refusal became illegal there.7
What the laws do
Massachusetts banned source-of-income discrimination in 1971. The pace picked up sharply in the last decade: as of September 2022, more than 57% of voucher households — about 1.3 million — lived in a place with some SOI protection, up from roughly 34%.5
Four findings, in order of how directly they measure what a renter would care about.
Utilization. Lance Freeman’s 2012 difference-in-differences analysis of 16 jurisdictions covering 48 housing authorities found voucher utilization rose 7 to 12 percentage points after SOI adoption, compared with adjacent jurisdictions without laws. Freeman himself flagged the limit: utilization is an agency performance metric, and an authority can improve it by managing how many vouchers it issues against expected turnover. It is not a direct measure of whether a household found a home.3
Where people end up. Freeman and Li later found modest reductions in the neighborhood poverty rates where voucher holders lived.3 Ellen, O’Regan and Harwood tracked the same families’ moves before and after enactment and found more upwardly mobile moves — but the effect only reached statistical significance three years after passage. Teles and Su, using a panel event study, found more voucher families with children moving into low-poverty neighborhoods, significant at about five years, with larger gains for Black families.4
Lease-up. The most direct measure — did the voucher get used before it expired — has the weakest evidence. Ellen, O’Regan and Strochak’s analysis of 85,000 voucher holders at 433 housing authorities found success rates modestly higher in states with SOI protections, with larger benefits for Black and Latino voucher holders. The authors describe this as descriptive and suggestive, not causal, and we’ll report it that way.2
Strong laws versus weak ones. Here the evidence cuts against the intuitive policy move. Teles and Su tested whether ordinances coded as “stronger” — bigger civil damages, criminal penalties, fewer exemptions — produced better location outcomes than weaker ones. They found no evidence that they did.4 The implication is uncomfortable for everyone: writing a tougher ordinance is the cheap part.
The gap nobody has measured
HUD’s own 2024 review of this literature is blunt that enforcement is close to a void in the research. Fair housing enforcement is complaint-driven, which means detection depends on the person least able to afford the delay — someone with a voucher clock ticking, who gains nothing personally from filing, and who often cannot tell whether the landlord who stopped answering was breaking the law or just busy. The National Fair Housing Alliance counted 2,395 source-of-income complaints nationally in 2022, against a covered population of 1.3 million voucher households. No one has compiled data on enforcement actions, testing capacity, or resources dedicated to SOI enforcement across jurisdictions.5
There is one encouraging line of work on the other side of the equation. Landlord refusal is not purely bias: interviews with 127 landlords in three cities found decisions driven by finances, past experience with tenants, and — repeatedly — frustration with housing authority inspections, timelines, and unresponsiveness.5 Oregon’s 2013 SOI law responded to that directly, pairing the ban with a landlord damage-mitigation fund and inspection-timeliness requirements, which is credited with winning landlord support.5 And a 2024 experiment in Pittsburgh found landlords who received asset-framed messaging about housing stability were significantly more likely to say they would accept vouchers than a control group.8 Said intentions are not signed leases. It is still the rare finding in this literature that suggests a lever other than a penalty.
What this means in Phoenix
Arizona has no statewide source-of-income protection. Phoenix and Tucson each passed one, and both survived a fight over whether cities may do so at all: the state Attorney General’s office concluded in December 2022 that Tucson’s ordinance violated state law and must be repealed, and the incoming Attorney General reversed that opinion in March 2023, clearing both.6 The Phoenix ordinance amended Chapter 18 of the city code and took effect in April 2023. As of today, the city’s Equal Opportunity Department lists source of income alongside race, disability and familial status as a protected basis in housing.6
If a Phoenix landlord turned you away because of a voucher: you can file a housing discrimination complaint with the City of Phoenix Equal Opportunity Department, and you have one year from the incident to do it. Call 602-262-7486 or email contact.eod@phoenix.gov. Before you call, write down the date, the property address, who you spoke to, and what they said — and screenshot the listing if the ad itself says “no Section 8.” Advertising language is the easiest kind of violation to prove. Outside Phoenix and Tucson city limits, Arizona law does not protect you, and a fair housing organization or legal aid is the better first call — our guide to free legal help in Phoenix lists them. If your housing is the emergency rather than the search, start at Find resources or call 211.
The honest summary: banning voucher discrimination is worth doing, and the research supports doing it. It is also a slow, partial fix that takes three to five years to show up in where families live, does not appear to work better when the penalties are bigger, and has never been paired with a serious measurement of whether anyone is enforcing it. A city that passes the ordinance and stops there has bought a real but small effect. The unglamorous follow-through — testing programs, faster inspections, landlord outreach, someone whose job is to notice the ads — is the part the evidence keeps pointing at and nobody has studied.
Frequently asked questions
Is it legal for a landlord to refuse a Section 8 voucher?
It depends entirely on where the apartment is. Voucher holders are not a protected class under the federal Fair Housing Act, so in most of the United States a landlord may legally refuse a voucher. Arizona has no statewide protection, but Phoenix and Tucson have each passed source-of-income ordinances. Inside Phoenix city limits, refusing to rent to someone because they hold a housing voucher — or advertising “no Section 8” — violates Chapter 18 of the Phoenix City Code.
How often do landlords actually refuse vouchers?
In HUD’s five-metro paired-testing study, refusal ranged from 15% of landlords in Washington, DC to 78% in Fort Worth, Texas. In every city except DC, refusal was more common in low-poverty neighborhoods — 85% of landlords tested in low-poverty Fort Worth neighborhoods and 82% in low-poverty Philadelphia neighborhoods turned vouchers away.
Do source-of-income laws work?
Yes, modestly and slowly. Voucher utilization rose 7 to 12 percentage points after jurisdictions adopted the laws. Effects on where voucher holders actually live took about three years to reach statistical significance in one study and about five in another. None of the evidence is a randomized trial. And one analysis found no evidence that laws with stronger penalties outperformed weaker ones, which points at enforcement rather than statutory text.
What can I do if a Phoenix landlord refuses my voucher?
File a housing discrimination complaint with the City of Phoenix Equal Opportunity Department within one year of the incident — 602-262-7486 or contact.eod@phoenix.gov. Document the date, address, name, and exact words, and screenshot the listing if the ad says “no Section 8.”
Sources & footnotes
- Cunningham M, Galvez M, Aranda CL, Santos R, Wissoker D, Oneto A, Pitingolo R, Crawford J, A Pilot Study of Landlord Acceptance of Housing Choice Vouchers, U.S. Department of Housing and Urban Development, Office of Policy Development and Research (2018); Urban Institute project page. Five metro areas: Fort Worth, Los Angeles, Newark, Philadelphia, and Washington, DC. Figures used here are as reported in HUD’s own 2024 review of the study (note 5): “whereas only 15 percent of landlords in Washington, D.C. rejected vouchers, 78 percent of landlords in Fort Worth, Texas, did”; “85 percent of landlords in low-poverty Fort Worth neighborhoods and 82 percent in low-poverty Philadelphia, Pennsylvania, neighborhoods rejected vouchers”; and an additional 15 percent in DC and 10 percent in Philadelphia gave unsure or conditional answers. Study scale (341,000 advertisements screened, 8,735 voucher-affordable units identified, ~3,780 landlords contacted, 4,000+ tests) is as described in the study and its published summaries. Per-city rates for Los Angeles, Philadelphia and Newark circulate widely in secondary coverage; we have quoted only the figures we could confirm in a HUD primary document.
- Finkel M, Buron L, Study on Section 8 Voucher Success Rates, Volume I, Abt Associates for HUD (2001) — about 69 percent of a sample of 2,600 voucher holders across 48 public housing authorities leased up within the allowed search time in 2000. Ellen IG, O’Regan KM, Strochak S, “Race, Space, and Take up: Explaining Housing Voucher Lease Up Rates,” Journal of Housing Economics 63 (2024) — HUD administrative data on 85,000 voucher holders at 433 housing authorities, success rate about 60 percent by 2019; success rates modestly higher in states with SOI protections, with larger benefits for Black and Latino voucher holders. Both figures and the authors’ own “descriptive” and “suggestive” characterization are quoted from the HUD review at note 5.
- Freeman L, “The Impact of Source of Income Laws on Voucher Utilization,” Housing Policy Debate 22(2):297–319 (2012) — difference-in-differences across 16 jurisdictions covering 48 housing authorities; utilization rose 7 to 12 percentage points after adoption relative to adjacent non-adopting jurisdictions. Freeman L, Li Y, “Do Source of Income Anti-Discrimination Laws Facilitate Access to Less Disadvantaged Neighborhoods?” Housing Studies 29(1):88–107 (2014) — modest reductions in neighborhood poverty and in the share of neighborhood residents who are people of color. The caveat that utilization is an agency-level performance measure rather than a household outcome is Freeman’s own, as reported in note 5.
- Ellen IG, O’Regan KM, Harwood KWH, “Advancing Choice in the Housing Choice Voucher Program: Source of Income Protections and Locational Outcomes,” Housing Policy Debate 33(4):941–962 (2022) — difference-in-differences on HUD administrative data comparing the same households’ moves before and after enactment; effects became statistically significant three years after passage, with larger improvements for households starting in the highest-poverty tracts. Teles D, Su Y, Source of Income Protections and Access to Low Poverty Neighborhoods, Urban Institute (2022) — panel event study; effect on families with children moving to low-poverty neighborhoods significant at about five years, larger for Black families than White or Latino families; no evidence that laws coded as “stronger” outperformed “weaker” ones. Both as summarized in note 5.
- Galvez M, Knudsen B, “Discrimination Against Voucher Holders and the Laws to Prevent It: Reviewing the Evidence on Source of Income Discrimination,” Cityscape 26(2):145–161 (2024), HUD Office of Policy Development and Research. This peer-reviewed review is the umbrella source for this brief and is the document we verified each secondary figure against. Program scale (about 2.3 million households, 2,400+ housing authorities); coverage (Knudsen 2022: more than 57 percent of voucher households, 1.3 million, across 16 states, DC, 21 counties and 85 local jurisdictions as of September 2022, up from roughly 34 percent); Massachusetts 1971 as the first SOI law; Garboden et al. (2018) interviews with 127 landlords in Baltimore, Cleveland and Dallas on financial, experiential and housing-authority drivers of refusal; Oregon’s 2013 law pairing the ban with a damage-mitigation fund and inspection-timeliness requirements (Galvez et al. 2020); 2,395 source-of-income complaints nationally in 2022 (National Fair Housing Alliance, 2023 Fair Housing Trends Report); and the authors’ statement that “very little is known about enforcing SOI laws” and that no comprehensive data exist on complaint frequency, enforcement actions, or resources. All causal estimates cited in this brief are quasi-experimental; the review identifies no randomized evidence on SOI laws.
- City of Phoenix Equal Opportunity Department, “Discrimination in Housing” — checked live September 3, 2026: Chapter 18, Article III of the Phoenix City Code makes it unlawful to discriminate in housing because of “race, color, religion, sex, disability, national origin, familial status, sexual orientation, gender identity, gender expression, or source of income”; complaints must be filed within one year of the incident; housing intake line (602) 262-7486, contact.eod@phoenix.gov, plus an online complaint form. Ordinance history: the Phoenix City Council voted 8–1 on March 1, 2023 to amend the code, effective the following month (Arizona Mirror, March 7, 2023; Arizona State Law Journal, April 18, 2023). Tucson passed its ordinance in September 2022 (NLIHC; City of Tucson SOI page); Attorney General Brnovich’s office concluded in December 2022 that it violated state law, and Attorney General Mayes reversed that opinion in March 2023. Arizona has no statewide source-of-income protection.
- Unlock NYC, Neighbors Together, Anti-Eviction Mapping Project, and Housing Data Coalition, An Illusion of Choice: How Source of Income Discrimination and Voucher Policies Perpetuate Housing Inequality (2022). Findings used here — that ghosting has become more common over time and that landlords apply minimum credit-score or income requirements to exclude voucher holders while remaining nominally compliant with New York’s SOI law — are as characterized in the HUD review at note 5.
- Ortiz SE, Fenelon A, Yousef C, “Exposing Pittsburgh Landlords to Asset-Framing Narratives: An Experiment to Increase Housing Voucher Participation,” Health Affairs 43(2):287–296 (2024). As summarized in note 5: landlords who received information about the importance of housing stability and access to high-opportunity neighborhoods were significantly more likely to say they would accept vouchers than a control group. The outcome is stated intention, not observed lease-up; full text is paywalled and we report it as the review characterizes it.