How many forced marriages exist in Germany?

The project focuses on forced marriage as a social phenomenon that is empirically difficult to capture due to its hidden, sensitive, and often stigmatized nature. Beyond its legal dimension, forced marriage constitutes a serious human rights violation and a form of gender-based violence that is embedded in social relations, institutional contexts, and power structures.

Germany currently lacks a unified and systematic approach to monitoring and addressing forced marriages, which makes it difficult to assess the true scope of the problem. To date, available evidence is largely limited to police and court statistics, a single nationwide report commissioned by the Federal Ministry for Family Affairs, Senior Citizens, Women and Youth (BMFSFJ, from 2024 – BMBFSFJ) in 2011, and fragmented data from individual support organizations and regional initiatives. As a result, forced marriages remain profoundly undermeasured at the national level.

At the same time, both governmental and non-governmental actors depend on credible and comprehensive data to inform policy decisions, develop effective prevention strategies, and ensure adequate support and protection for affected individuals.

This project addresses this gap by providing the first population-based estimate of the prevalence of forced marriages in Germany. Beyond producing an empirical benchmark, it establishes a methodological foundation for future research, including more targeted sampling strategies, and underscores the urgent need to strengthen support structures for affected individuals.

The main goal of the project is to provide a robust empirical estimate of the number of people living in forced marriages in Germany based on general population survey data. By generating reliable evidence on the scale of forced marriages, the project seeks to support evidence-based debates on prevention, victim support, and policy responses, and to inform future research and monitoring efforts

This study is the first to apply the Network Scale-Up Method (NSUM) to estimate the number of people living in forced marriages in Germany. NSUM is an established indirect estimation technique used to study hidden or hard-to-reach populations (that is, groups that are difficult to identify through direct surveys because of stigma, fear, or legal and social risks). Instead of asking individuals whether they themselves are affected, NSUM infers the size of a hidden population from information about people’s social networks, using a representative sample of the general population. The basic idea behind NSUM is simple: people can reveal facts about their social environment that they are reluctant to disclose about themselves.

Our analysis draws on wave 12 of the DeZIM.panel. Data were collected from September 16, 2024 to November 25, 2024. Of the 7,236 individuals invited, 3,792 respondents completed the survey. DeZIM.panel is representative of the population aged 18 to 67 living in Germany, with targeted oversampling of individuals with migration backgrounds; survey weights were applied to restore population representativeness.

The estimation procedure can be summarized in four main steps:

1. Collection of aggregated relational data. Respondents are asked how many people they know within certain social groups, using questions such as “How many X do you know?”, where X refers to:

• multiple known populations (e.g. groups with reliable official statistics), and
• the unknown population of interest (individuals living in forced marriages).

2. Using responses about known populations and their true population sizes, the model infers each respondent’s network size.

3. Respondents’ answers about acquaintances in forced marriages are combined with their estimated network sizes to derive the proportion of individuals in forced marriages within each respondent’s social network.

4. Finally, under the core NSUM assumption that social networks, on average, reflect the composition of the general population, this proportion can be extrapolated to the population level. We employ one basic Frequentist NSUM model and three Bayesian models: (1) an Overdispersed model (Zheng et al., 2006), (2) a Correlated NSUM model without demographic controls, and (3) a Correlated NSUM model with demographic controls, following (Laga et al., 2023). The Bayesian framework allows us to explicitly account for uncertainty, overdispersion in count data, and systematic network distortions. This results in a posterior distribution for the number of individuals living in forced marriages in Germany. 

A marriage is considered forced when one or both individuals are compelled – through coercion, threats, or the risk of significant harm – to enter into a formal or informal union, whether through a legal process or a religious or social ceremony. Individuals may feel they have no real choice regarding when or whom they marry or fear expressing their true wishes due to potential negative consequences.