Against the backdrop of sustained global fertility decline, raising fertility intentions among people of reproductive age has become a central public policy concern. Existing research has largely examined their determinants at the macro-structural and individual levels, while meso-level explanations from a social network perspective remain underdeveloped. Yet fertility decision-making is socially embedded and shaped by surrounding others within social networks, giving rise to group effects in childbearing. Drawing on neighborhood social networks, this study incorporates community-level fertility into the analysis of individual fertility intentions and systematically examines the existence, heterogeneity, and mediating pathways of such group effects. It thereby extends fertility-intention research from a social network perspective and provides evidence for improving fertility support policies.

Using data from the 2021 Chinese General Social Survey (CGSS), the study focuses on childless individuals of reproductive age. Multiple linear regression and logistic regression models estimate the effects of community-level fertility on fertility intentions. Subgroup regressions examine heterogeneity by gender, educational attainment, and age, together with a gender-by-education cross-classified subgroup analysis. A bias-corrected bootstrap mediation analysis tests the psychological pathways of group effect, with endorsement of childbearing values and perceived barriers to childbearing as mediators.

The results reveal significant group effects on fertility intentions. Individuals living in communities with higher fertility levels report a larger ideal number of children, a lower likelihood of preferring childlessness, and a stronger intention to have a second child. These findings remain robust after an instrumental variables approach is used to mitigate potential endogeneity bias. The effects also have clear boundary conditions. Community-level fertility significantly influences men but not women, individuals with lower but not higher educational attainment, and those aged 18-39 but not those aged 40-49. The gender-by-education analysis further shows that among highly educated women, the effect is statistically insignificant and even tends to be negative. The mediation analysis indicates that community-level fertility positively affects fertility intentions indirectly by strengthening endorsement of childbearing values and reducing perceived barriers to childbearing.

This study identifies the boundary conditions and psychological pathways of group effects and makes three theoretical contributions. First, it introduces a social network perspective into fertility-intention research and shows that neighborhood norms have not disappeared but persist heterogeneously. Second, it verifies two psychological pathways---endorsement of childbearing values and perceived barriers to childbearing---addressing the literature's emphasis on effect identification over mechanism analysis. Third, systematic heterogeneity analysis identifies gender-, education-, and age-specific boundary conditions, especially the absence of detectable group effects among highly educated women, helping explain the structural constraints they face in childbearing.

These findings have important policy implications. Group effects are substantially suppressed among women, particularly highly educated women, who bear greater physiological costs, career opportunity costs, and day-to-day caregiving burdens. Fertility support therefore cannot rely solely on fostering a supportive social climate or collective mobilization. Policies should instead address the practical constraints faced by women, who bear the primary costs of childbearing. They should also promote gender equality within the household, establish incentives for paternal involvement in childcare, and facilitate the redistribution of caregiving responsibilities within families. These measures may constitute a key policy leverage point for supporting childbearing.