Background Personal characteristics, income, social capital, and immigration policies influence immigrants’ social integration (SI) in European Union (EU) countries. However, few studies have used machine learning (ML) to examine simultaneously the predicting significance of large number of variables (93 in total). Objectives The present study aimed at, first, assessing the level of SI in a random sample of non-EU new immigrants legally living in France; second, use a data-driven, exploratory, and a machine learning (ML) approach to determine the most important predictors of the participants’ SI score, three years after their residence permit approval. Methods The study was based on data from the government-led longitudinal survey (ELIPA-2). Specifically, the dataset used included 4,053 French new immigrants (age = 18-84, mean age = 34.25; 57% of male, 43% of females, 71% originate from Africa) who participate in the three waves (2019, 2020, 2022). It comprised a total of 93 predictor variables + the outcome variable. Data was analyzed using descriptive analyses, t-test and chi-square analyses, univariate Pearson's correlations, and multivariate ML regression analysis. Results The SI level was above the measurement scale's middle point (SI mean = 47.10; 10-82 scale). Positive predictors were income, highest study degree/diploma, general health, French knowledge, documents’ legality and accommodation quality upon arrival in France, sense of belonging to home country before emigration, and life satisfaction in France. Negative predictors were respondents’ age, length of presence in France, resident permit's nature and duration, and the country of origin. Conclusion and implications First, these findings underscore the heightened integration challenges faced in France by newly arrived immigrants originating from countries characterized by markedly different sociocultural values and social norms. Such cultural distance may complicate processes of social participation, identity negotiation, and access to institutional resources, thereby impeding successful settlement trajectories. Second, the results indicate that individuals who migrate in the context of traumatic experiences—such as armed conflict, gender-based violence, or persecution related to sexual orientation—encounter additional and often more complex barriers to integration in the host society. The cumulative effects of pre-migration trauma, displacement-related stressors, and post-migration adversities may substantially hinder psychological adjustment, social inclusion, and economic participation. From a policy perspective, these findings highlight the need for differentiated and trauma-informed integration strategies. Rather than adopting a uniform approach, public policies should incorporate targeted, culturally responsive, and psychosocially informed programs tailored to the specific profiles and vulnerabilities of these distinct groups of newcomers.
Cruz et al. (Mon,) studied this question.