Ambivalence and transnational intergenerational solidarity: the perspective of highly educated Portuguese women emigrant daughters

The study investigates how highly educated Portuguese adult daughters living abroad perceive intergenerational solidarity with parents who remain in Portugal, emphasizing the role of ambivalence as a transversal dynamic shaping solidarity dimensions. Building on developments in the intergenerational solidarity-conflict model, the authors integrate ambivalence to understand ties during transitional periods such as migration. Portugal provides a pertinent context given its high emigration rates, increasing shares of highly skilled migrants, and a familialist welfare regime with traditional gendered care norms. The purpose is to construct an ambivalence-informed typology of transnational parent–daughter relationships and to examine how solidarity dimensions co-occur with perceived ambivalence, acculturation, sociocultural adaptation, and social well-being.

The theoretical framework draws on Bengtson’s multidimensional model of intergenerational solidarity (affectual, consensual, functional, associational, structural, normative) later expanded with “systematic” solidarity (Cavallotti et al., 2017) and extended to include conflict and ambivalence. Prior typology work has identified patterns of parent–adult child relations in general and among migrants, but rarely integrates ambivalence as a sensitizing construct. Ambivalence acknowledges simultaneous positive and negative dynamics and may catalyze change in family relations. Gendered norms are salient: daughters often experience more ambivalence than sons, especially when caregiving demands rise with parental aging and when exits (e.g., sibling support) are limited. Structural contexts (welfare regimes, labor market participation) shape ambivalence by constraining negotiation within families. In migration contexts, acculturation and sociocultural adaptation relate to solidarity and ambivalence: lower adaptation often links to higher ambivalence and poorer solidarity, though adjustment can also enhance cohesion. Social well-being, particularly social cohesion and networks in the host context, may buffer or correlate with solidarity patterns. This study adopts a multilevel perspective linking individual experience, family ties, and macro contexts to analyze ambivalence–solidarity typologies among highly educated Portuguese women abroad.

Design and sample: Cross-sectional online survey of N=248 highly educated Portuguese adult daughters (≥18 years; minimum bachelor’s degree) living transnationally with parents in Portugal. Non-probabilistic recruitment via social networks, media, institutional pages, and municipal/consular mailing lists. Data collected over four months in spring 2019 via Qualtrics. Ethics approval obtained; informed consent secured. Measures: Intergenerational solidarity measured across five traditional dimensions—normative (11 items, α=0.80), associational (7 items, α=0.88), affectual (4 items, α=0.72), consensual (4 items, α=0.68), functional (2 items, α=0.70)—plus systematic solidarity (5 items, α=0.60). Psychological ambivalence assessed with a 9-item scale (Michels et al., 2011; α=0.85). Adaptation and acculturation: Brief Sociocultural Adaptation Scale (12 items, α=0.80) and Brief Acculturation Orientation Scale (6 items, α=0.61). Social well-being: social cohesion (5 items, α=0.79) and perceived satisfaction with social connections (4 items, α=0.85). Responses used 5- or 7-point Likert scales; higher scores indicate more of the construct. Procedure and analysis: Descriptive statistics computed for all variables. Dimension reduction via principal component analysis (PCA) on z-standardized items from solidarity dimensions and ambivalence; factor retention by Kaiser criterion and scree plot; promax rotation. Sequential (tandem) clustering via k-means on extracted components; cluster quality assessed by within-cluster sums of squares and silhouette width. Visualization through PCA loadings plot and parallel coordinate plots of standardized cluster means. Group differences examined using MANOVA (Pillai’s trace due to unequal cell sizes and covariance differences) and Tukey–Kramer post hoc tests for unequal n. Software: JASP 0.16.2; JAMOVI 2.3.3. Ethical compliance followed APA guidelines; confidentiality and anonymity assured.

Descriptives: Mean age 36.16 (SD=7.06), mean length of residence abroad 7.82 years (SD=5.99). Most resided in Europe (74%) and urban areas (76%). Education: 43% bachelor, 42% master, 16% PhD. Employment: 86% employed. Solidarity means (range 1–5 unless noted): normative 4.19 (0.42), associational 3.95 (0.52), consensual 3.15 (0.64), functional 4.02 (1.27), affectual 3.60 (0.60), systematic 2.99 (0.66). Psychological ambivalence (1–7): 3.20 (1.29). Acculturation orientation 5.27 (0.81); sociocultural adaptation 5.21 (0.88); social cohesion 5.03 (0.66); social connections 4.53 (1.31). PCA and clustering: Two principal components retained, explaining 50.8% variance. Dimension 1 loaded positively on normative, associational, consensual; ambivalence loaded negatively with these; functional showed weak correlation. Dimension 2 characterized by ambivalence and functional (positively) vs affectual (negatively). Three clusters identified (average silhouette ≈0.20):

• Cluster 1 (Type 1: Autonomous with affection and low ambivalence), n=97 (39.9%): higher affectual; lower than average normative, associational, consensual, functional, and systematic; low ambivalence.
• Cluster 2 (Type 2: Ambivalent functional ties with low affection), n=46 (18.9%): highest ambivalence; above-average functional; low normative, associational, consensual, and very low affectual; moderate systematic.
• Cluster 3 (Type 3: Low ambivalence with strong cohesion), n=100 (41.2%): highest across all solidarity dimensions (especially normative, associational, consensual, systematic), with low ambivalence; functional around average. Group differences (MANOVA): Significant overall effect of clustering, Pillai’s V=1.351, F(36,438)=25.326, η²=0.68, p<0.001. Tukey–Kramer contrasts:
• Solidarity dimensions: Type 3>Type 1=Type 2 for normative and associational (p<0.001); consensual: Type 3>Type 1>Type 2 (p<0.001). Affectual: Type 3>Type 1>Type 2 (p<0.001). Systematic: Type 3>Type 2>Type 1 (p<0.001). Functional: Type 2>Type 1 (p<0.01); Type 2≈Type 3; Type 1≈Type 3.
• Ambivalence: Type 2>Type 1=Type 3 (p<0.001).
• Social well-being: Social cohesion higher in Type 3 than Type 2 (p<0.05); social connections did not differ.
• Acculturation: Acculturation orientation higher in Type 3 than Types 1 and 2 (p<0.001); sociocultural adaptation did not differ significantly.
• Socio-demographics: Type 2 had older mean age than Type 3 (p<0.05); more likely single and economically inactive/unemployed. Longer residence abroad associated with Type 1 vs Type 3 (p<0.05). Distribution patterns suggest Type 1 more prevalent in de-familialized welfare contexts (e.g., Sweden, Denmark, Netherlands). Interpretation: High ambivalence was linked to financial exchanges and low affect/consensus (Type 2), especially among older, single, and non-employed daughters. Low ambivalence was common (~80%) but manifested either as strong cohesive solidarity across dimensions (Type 3) or as autonomous ties with high affect but low other solidarities (Type 1). Better integration (higher social cohesion and acculturation orientation) was associated with stronger cohesive ties (Type 3).

The findings demonstrate that ambivalence differentially qualifies intergenerational solidarity dimensions in transnational mother–daughter relations. Rather than a single continuum, solidarity and conflict/ambivalence co-occur in patterned ways. Incorporating ambivalence revealed three meaningful relationship types. High ambivalence aligned with functional exchanges and diminished affect and consensus, indicating potential strain where financial dependence or obligations exist. Low ambivalence split into two distinct forms: strong-cohesion ties with high solidarity across domains and autonomous affectionate ties marked by emotional closeness but low contact, exchange, and normative/systematic engagement. Macro contexts and integration in the host society matter: higher social cohesion and stronger orientation toward the host culture were associated with cohesive, low-ambivalence ties, while longer residence and de-familialized contexts related to autonomous affectionate ties. These results address the research questions by showing how ambivalence interrelates with solidarity dimensions to form types, how acculturation relates to these types, and which socio-demographic characteristics differentiate them. They underscore a spillover effect whereby successful integration corresponds to higher-quality intergenerational relations—either interdependent or independent but affectively positive and low in ambivalence.

This study advances research on transnational intergenerational relations by integrating ambivalence into a typological analysis of solidarity among highly educated Portuguese emigrant daughters. Three types emerged: autonomous with affection and low ambivalence; ambivalent functional ties with low affection; and low ambivalence with strong cohesion. High ambivalence clustered with functional exchanges and lower affect/consensus, especially among older, single, and economically inactive/unemployed daughters. Low-ambivalence ties bifurcated into cohesive high-solidarity patterns associated with stronger acculturation orientation and social cohesion, and autonomous affectionate ties associated with longer residence abroad and de-familialized host contexts. The study suggests a spillover effect of successful integration into improved family relations. Future research should use longitudinal designs to track transitions between types over the life course, incorporate gender comparisons, and include more detailed measures of dependency structures and parental health, to inform policy on balancing formal and informal care and understanding how ambivalence regulates family dynamics over time.

• Cross-sectional design prevents causal inference and understanding of temporal dynamics or transitions between relationship types.
• Non-probabilistic, self-selected online sample of highly educated Portuguese women limits generalizability and may introduce selection bias.
• Sample size (N=248; with small cluster n for Type 2) may reduce power to detect less common patterns, such as high ambivalence coexisting with high solidarity.
• Reliance on self-report measures may introduce reporting biases.
• Structural solidarity was not used due to uniformly high geographic distance, potentially omitting a contextual dimension.
• Acculturation orientation scale exhibited modest internal consistency (α≈0.61).
• Findings are specific to Portuguese cultural and welfare-state contexts and may not generalize across migrant groups.