Two to Tango: Are Spouses' Perception of Women's Empowerment Associated with Concordance on Fertility Desires?

Technical Summary: Spousal Concordance on Fertility Desires in Pakistan

Problem Statement
Traditional family planning research in Pakistan has predominantly focused on women's perspectives, often overlooking the dyadic nature of reproductive decision-making. While fertility decisions typically involve joint negotiation, large-scale surveys frequently rely on women's reports of their partners' desires, which may not accurately reflect male preferences. This gap is critical because spousal discordance on fertility desires is linked to lower contraceptive use and is driven by gendered social norms and power imbalances. Despite a decline in Pakistan's fertility rate from 7.0 in 1990 to 3.6 in 2017-18, progress has stalled, with modern contraceptive prevalence remaining stagnant at 25%. The study addresses the lack of couple-level data regarding how spouses' perceptions of women's empowerment influence concordance on fertility desires.

Methodology
The study utilized cross-sectional data from the Pakistan Demographic and Health Survey (PDHS) 2017-18, analyzing a sample of 3,027 currently married couples. Couples where either spouse was sterilized or infecund were excluded.

• Outcome Variable: Spousal concordance on fertility desires was categorized into four groups based on separate responses from husbands and wives to the question, "Would you like to have (a/another child, or would you prefer not to have any (more) children?"):
  1. Concordant (want more)
  2. Husband+ discordant (husband wants more, wife does not)
  3. Wife+ discordant (wife wants more, husband does not)
  4. Concordant (no more)
• Independent Variables: The study operationalized women's empowerment using Connell's Theory of Gender and Power, focusing on three structures: sexual division of power (household decision-making), sexual division of labor (economic/educational factors), and cathexis (emotional/relational dimensions). Key constructs included household decision-making authority and attitudes toward Intimate Partner Violence (IPV), reported by both spouses.
• Data Analysis:
  • Concordance Measurement: The Kappa coefficient was used to assess the level of agreement between spouses.
  • Variable Construction: Principal Component Analysis (PCA) was applied to assess the internal consistency and dimensionality of empowerment items. Cronbach's alpha (>0.7) and eigenvalues (>1.0) confirmed high internal consistency. Items were summed to generate categorical empowerment scores (e.g., low, moderate, high decision-making authority).
  • Regression Modeling: Multinomial logistic regressions were performed to estimate Relative Risk Ratios (RRR). The models assessed the association between empowerment variables and spousal concordance, adjusting for covariates. Interaction terms were included to test if husbands' perceptions of women's empowerment moderated the association between wives' self-perceived empowerment and concordance.
  • Software: Analyses were conducted using Stata 12 with svy commands to account for the complex two-stage cluster sampling design.

Key Results

• Level of Concordance: The study found a moderate level of spousal agreement on fertility desires, with a Kappa value of 0.51.
• Empowerment and Discordance:
  • Wife's Self-Perception: When a wife perceived herself as having high decision-making authority, she was more than twice as likely (RRR=2.06, 95% CI: 1.20–3.55) to be in a "wife-only discordant" state (wanting more children while the husband wanted none) compared to couples where both wanted more children.
  • Husband's Perception: Conversely, when the husband perceived his wife as having high decision-making authority, the likelihood of "wife-only discordance" decreased by 62% (RRR=0.38, 95% CI: 0.21–0.66).
  • IPV Attitudes: High husband intolerance toward IPV was associated with a 2.48-fold increase in "husband-only discordance."
  • Negotiation: A wife's ability to negotiate sex increased the likelihood of concordance on wanting "no more" children by 2.09-fold.
• Interaction Effects: When both spouses perceived the wife as having moderate or high decision-making authority, the likelihood of neither spouse wanting more children was significantly lower (approximately 70-80% lower) compared to couples where both wanted more children.
• Null Findings: A central finding was that 23 of the 27 adjusted regression coefficients were not statistically significant, indicating that the empowerment constructs examined largely failed to predict variation in spousal concordance outcomes in this specific sample.

Significance and Claims
The paper claims to provide nationally representative evidence on the dyadic nature of fertility desires in Pakistan, identifying household decision-making authority as the most consistently associated dimension of empowerment. The study highlights that while women's empowerment is a critical factor, its relationship with fertility concordance is complex and not linear.

The authors argue that the lack of significant associations for most empowerment constructs suggests that in settings with strong pronatalist norms, fertility preferences may be shaped more by shared social expectations than by within-couple power dynamics. The study concludes that effective family planning interventions must move beyond individual empowerment to address couple-level dynamics, shared social norms, and spousal communication. It advocates for policies that encourage men's inclusion in family planning programs and couple counseling to facilitate reproductive decision-making.

Limitations Noted by the Authors
The authors explicitly note that the PDHS data lacks information on the quality of spousal communication, relationship satisfaction, and community-level gender norms. They also acknowledge that the categorical operationalization of empowerment scores may have reduced statistical sensitivity and that social desirability bias could lead to underreporting of fertility desires. The study calls for future research using purpose-built instruments, longitudinal designs, and Actor-Partner Interdependence Models (APIM) to better disentangle these dynamics.