Housing discrimination is usually visible to researchers before it is provable to the person who experienced it. A rejected applicant may see only one unanswered message, one cancelled viewing, or one additional document demand. Research becomes powerful when many similar moments are compared.
This article reviews what can be known from housing research in Germany and comparable European contexts. It does not advise an individual tenant or applicant through a case. Shelter & Stability Network covers the practical evidence problem during a housing search, and Equal Voices Initiative covers discrimination response more broadly.
The Concept: Unequal Treatment in a Tight Market
Housing discrimination means that people are treated worse in access to housing, contract conditions, communication, or tenancy relations because of protected or socially marked characteristics such as racialization, name, perceived religion, nationality, family status, disability, gender, or age. In German public debate, this often appears around surnames, accents, headscarves, family size, benefit receipt, or assumptions about origin.
A tight rental market complicates interpretation. Many people receive no answer because demand is high. A landlord may choose one applicant from dozens. Scarcity can hide discrimination because rejection looks ordinary. It can also intensify discrimination because landlords or agents have more room to filter applicants without explanation.
The research question is therefore not “Can every rejected applicant prove discrimination?” Usually they cannot. The research question is “When applicants are comparable in relevant ways, do some groups receive systematically worse treatment?” That is where the evidence is strongest.
What’s Known: Audit and Correspondence Studies
Audit and correspondence studies are the core method family for measuring unequal treatment in housing access. In a correspondence study, researchers send equivalent applications or inquiries that differ in one signal, such as a name associated with different ethnic or religious backgrounds. In an audit study, trained testers may contact providers or attend viewings using matched profiles.
The strength of this method is comparison. If two applications are similar in income, employment, household size, message quality, and timing, but one receives fewer replies after a name or other signal changes, unequal treatment becomes measurable as a pattern. No single message proves the whole pattern. The design does.
These studies in Germany and other European housing markets have repeatedly found that signals connected to migration, racialization, or religion can reduce the chance of receiving a reply or viewing invitation. The exact size of the difference varies by city, market segment, method, and signal used. The stable finding is not a precise universal number. It is that unequal treatment can be detected even at the first contact stage, before a landlord has met the person.
The contested parts are also important. Names do not measure identity cleanly. A name may be read as ethnicity, religion, nationality, class, or language background at once. Audit designs usually simplify applicant profiles to isolate treatment, but real people are not simplified profiles. Studies can also miss discrimination that happens later, such as during viewings, contract negotiation, repairs, deposit disputes, or neighbourhood conflict.
What’s Known: Surveys and Lived Reports
Surveys ask people whether they have experienced discrimination in the housing market, how often, in what situations, and with what consequences. They can capture experiences that audit studies miss: repeated humiliation, self-exclusion from certain neighbourhoods, fear of calling a landlord after hearing a tone change, or the cumulative cost of being asked for extra documents.
Surveys also allow researchers to examine who reports discrimination and how it relates to income, household structure, language, disability, gender, and migration history. In Germany, broader data infrastructures such as the Mikrozensus or SOEP help describe housing conditions and social inequality, even when they are not designed specifically as discrimination reporting systems.
The limitation is that surveys depend on recognition, memory, wording, trust, and willingness to answer. Some people may not label unequal treatment as discrimination even when it fits the pattern. Others may be unsure because housing searches are already opaque. A non-response can be read as racism, class filtering, market overload, or all three; the applicant often receives no explanation.
This does not make survey evidence weak. It means survey evidence answers a different question: what people experience and perceive across their search, and how these experiences are distributed. It is especially useful when combined with audit evidence, interviews, and administrative context.
What’s Known: Administrative Data and Complaint Records
Administrative data can show parts of the housing system: social housing access, homelessness services, rent levels, neighbourhood composition, benefit receipt, and sometimes outcomes of formal complaints. Complaint records from anti-discrimination bodies or advice services can show which issues reach support systems and what patterns practitioners observe.
These sources are valuable, but they are not a complete measure of discrimination. Most discriminatory treatment in housing does not become an official complaint. People may not know where to report. They may need housing urgently and fear losing time. They may worry about retaliation, language barriers, legal costs, or being dismissed. They may also lack evidence because the interaction was a phone call, a viewing, or silence.
Official complaint numbers therefore usually underestimate discrimination. A low number can mean low reporting, not low harm. This is called a reporting gap: the distance between experiences that occur and experiences that enter formal records.
Administrative data has another limit. Categories are often built for service delivery, not for research on discrimination. They may record nationality but not racialization; household type but not whether a family was rejected because children were unwelcome; disability status in one system but not accessibility barriers in private rental search.
Individual Cases and Research Patterns Are Not the Same Thing
Research can show that a pattern exists without proving what happened in any single case. This distinction matters ethically and legally.
An individual applicant may have a strong sense that a rejection changed after a name, accent, headscarf, wheelchair, children, or benefit status became visible. That experience should not be dismissed simply because motive is hard to prove. At the same time, a research institute cannot infer the landlord’s reason from one incomplete interaction.
Patterns answer different questions: Are similar applicants treated differently? At which stage? For which signals? Under what market conditions? Which groups face the highest barriers? Which policies or practices might reduce unequal filtering?
Individual response requires a different lane. Shelter & Stability Network explains how housing applicants can compare requirements, preserve messages, and think about evidence during a search. Equal Voices Initiative helps with discrimination documentation and reporting choices. The Institute’s role is to clarify what the research can support in public reasoning.
What’s Contested: Mechanisms
Researchers often distinguish between several mechanisms, but in real markets they overlap.
Some unequal treatment may be direct prejudice: a provider does not want to rent to a particular group. Some may be statistical discrimination: a provider uses group stereotypes as shortcuts for assumed risk. Some may be institutional: application systems, language expectations, viewing schedules, document demands, or referral networks favour some applicants over others. Some may be economic filtering that interacts with discrimination, such as excluding people who receive public benefits or cannot produce documents common among long-settled tenants.
The mechanism matters because solutions differ. If discrimination happens at first contact, anonymous or standardized inquiry processes may help. If it happens through document demands, transparency about required documents matters. If it happens through informal networks, public listings and allocation rules matter. If it happens through inaccessible buildings and communication channels, disability access is not a side issue.
Research rarely identifies one mechanism perfectly. Strong evidence often comes from triangulation: audit studies show unequal treatment; surveys show lived patterns; interviews explain how applicants adapt; administrative data shows where exclusion accumulates.
What We Cannot Say From the Evidence Alone
We cannot say that every unanswered message is discrimination. We cannot say complaint numbers represent the true scale. We cannot say that one method captures the whole rental pathway. We cannot say that a finding from one city, price segment, or year automatically applies everywhere else.
We also cannot say that discrimination is only an attitude problem. Housing access is shaped by scarcity, wealth, inheritance, local policy, tenant protections, social housing availability, credit histories, language, disability access, and informal trust. Discrimination operates inside those structures, not outside them.
The responsible public claim is narrower and stronger: research using matched inquiries and related methods shows that unequal treatment in housing access is measurable; survey and advice evidence shows that affected people experience barriers beyond isolated rejection; official complaint numbers undercount the issue; and individual cases need their own support route.
That is enough to reject the idea that housing discrimination is only a collection of misunderstandings. It is also enough to keep us from overstating what a single dataset can prove.
