In times of political polarization, understanding the factors that shape political attitudes and behavior is crucial. Recent research and policy reports have emphasized the importance of psychosocial factors, pointing to possible links between polarization and ongoing crises in mental health and social belonging. The central aim of this doctoral thesis was to advance knowledge about how psychosocial factors influence political preference formation across the life course. To do so, I drew on large-scale longitudinal data spanning key developmental stages, from adolescence to midlife.
This thesis examined three important psychosocial domains: mental health, self-esteem, and interpersonal relationships, and their role in shaping political preferences in adulthood. Together, these factors were studied in relation to more than a dozen indicators of political attitudes and voting behavior. The project comprised three studies based on two longitudinal datasets from Norway and New Zealand—both multi-party political systems characterized by coalition governments. Although this is the dominant democratic model worldwide, political attitudes in multi-party contexts remain relatively understudied.
Each study employed distinct longitudinal designs, offering different advantages for examining how psychosocial factors and political preferences are interrelated. Papers 1 and 2 used Norwegian data following 2,215 participants across five waves spanning 28 years, beginning in adolescence (1992) and ending in midlife (2020). Paper 3 drew on annual dyadic data from 1,613 intimate partner couples in New Zealand, followed for 10 years (2013–2022). Paper 1 used latent class growth analysis to assess how internalizing problems (depressive symptoms and anxiety) and externalizing problems (conduct problems) from adolescence to midlife related to 12 political outcomes, including political orientation, trust, and voting behavior. Paper 2 applied latent growth curve modeling to three types of self-esteem—general, social, and close friendship—to examine to what extent their developmental trajectories are associated with polarization-related attitudes including social dominance orientation, opposition to immigration, and opposition to gender equality. Paper 3 used actor–partner interdependence random-intercept cross-lagged panel models to test whether changes in one partner’s political preferences predicted changes in the other partner’s preferences the following year.
The findings provide several key insights. First, the analysis of mental health trajectories identified three classes for internalizing problems and two for externalizing problems. Individuals in the classes with high levels of either showcased lower political and social trust in midlife. Furthermore, high internalizing problems were associated with more left-leaning orientations and attitudes, while those with high externalizing problems demonstrated more right-leaning populist tendencies. Second, the self-esteem analyses showed that low self-esteem in adolescence, as well as depressed self-esteem across three decades, were associated with stronger social dominance orientation and greater opposition to immigration and gender equality. These findings were consistent across all self-esteem domains and while controlling for adolescent conservatism and other key covariates. Third, the intimate partner study demonstrated that increases in one partner’s party support predicted increases in the other partner’s support the following year. These partner effects appeared for six political parties and, with one exception, were equally strong among women and men.
Taken together, the three studies lead to two overarching conclusions about psychosocial development and political preferences in adulthood. First, adolescence is a critical period: psychosocial challenges such as internalizing or externalizing problems or low self-esteem are associated with distinct political outcomes decades later, particularly when these challenges persist beyond adolescence. Second, political attitudes continue to evolve in adulthood. Intimate partners play a role in shaping these changes. In sum, this thesis demonstrates that political attitudes and behaviors are not fixed early in life but are shaped by long-term psychosocial trajectories. Drawing on decades of longitudinal data, it provides clearer insight into how mental health, self-esteem, and interpersonal relationships relate to political preferences, while also highlighting pathways that may contribute to political polarization.