The research, led by Karen Benjamin Guzzo of the University of North Carolina at Chapel Hill together with Alexandra VanBergen and Claire Kamp Dush of the University of Minnesota and Wendy D. Manning of Bowling Green State University, uses data from the National Couples’ Health and Time Study (NCHAT), a probability-based national survey of cohabiting and married American adults aged 20 to 60 fielded by the Gallup organization between September 2020 and April 2021. Crucially, for a subset of respondents both members of the couple completed the survey, producing true dyadic data rather than the proxy reports that have plagued earlier couple-level research. After restricting the sample to different-cisgender couples in which both partners were between 20 and 50 years old, not pregnant, and complete on key covariates, the analytical sample comprised 608 couples, meaning 608 men and 608 women whose answers could be compared directly.

The theoretical foundation of the study is deliberately hybrid. The authors apply the Narratives of the Future (NoF) framework, developed largely in Europe to explain post-recession fertility declines that conventional economic models could not account for, to the Traits–Desires–Intentions–Behavior (TDIB) model of fertility psychology. In the TDIB sequence, conscious and subconscious dispositions, or traits, shape desires; desires precede firmer intentions, which in turn inform behavior. The researchers argue that uncertainty, stress, and perceptions of current and future well-being belong squarely within the traits component, and because traits influence desires directly but intentions only indirectly, fertility desires, an often-overlooked stage at which ambivalence is most visible, should be the focus of attention. This matters, they contend, because childbearing is inherently future-oriented: people ask whether they can muster the resources, broadly defined, to parent in a way that meets the intensive parenting expectations now pervasive in American culture.

A second conceptual pillar is the life course perspective, with its emphasis on multidimensionality and linked lives. Fertility decisions are overwhelmingly made within couples, since fewer than one in seven U.S. births occurs outside a coresidential union, yet couple-level analyses with direct reports from both partners have been rare in the American context. The life course lens also prompted the team to widen the analytical aperture beyond money. Health concerns were arguably the defining stressor of the pandemic period, with unclear early information about transmission, differential exposure across populations, fears about infection during pregnancy, and restrictions that barred partners from prenatal appointments and deliveries. Relationship stress, too, had distinctive pandemic contours: some couples accelerated their commitment within quarantined “pods,” while others questioned their partnerships under the strain of confinement.

The study’s dependent variable at the couple level was constructed from each partner’s answer to a single question: “Would you, yourself, want a(nother) child?” Responses on a five-point scale from definitely yes to definitely no, with an explicit “not sure” option, were collapsed into a three-category joint outcome: both partners want a(nother) child, both do not want a(nother) child, or the couple disagrees or both are unsure. The independent variables were three pandemic-specific stress scales. Health stress averaged items on worry about oneself, one’s partner, and one’s family members getting coronavirus, with reliabilities of 0.89 and 0.87 for main respondents and partners respectively. Economic stress combined worry about money and finances, one’s job, and obtaining food and supplies. Relationship stress drew on items asking whether the pandemic made respondents question their relationship or expect a breakup, with one positively worded item reverse-coded. All three were measured on one-to-five scales.

The descriptive picture is striking in itself. Just over half of couples, 53.7 percent, agreed they did not want any or any more children, while 22.1 percent agreed they did want a(nother) child. The remaining 24.2 percent either disagreed with each other or were both unsure, and within that category a full 93 percent were couples who flatly disagreed rather than shared indecision. Average couple-level stress was highest for health, at 2.76, followed by economic stress at 2.40, with relationship stress lowest at 1.68. Roughly a quarter of American couples, in other words, did not know or did not agree on whether their family would grow.

The multivariable analyses, multinomial logistic regressions reporting relative risk ratios for the couple-level outcome and ordered logit models for individual desires, sharpened the story considerably. Economic stress emerged as the strongest and most consistent driver of couple-level discord. Each one-unit increase in average couple economic stress doubled the relative risk of disagreement or shared uncertainty compared with both partners wanting a child, with a relative risk ratio of 2.03, and raised it by roughly 75 percent compared with both partners agreeing they did not want a child, at 1.77. Health stress similarly increased the risk of discordance relative to shared wanting, with a relative risk ratio of 1.81. The effect of relationship stress was even larger: couples with higher relationship stress had 2.77 times the relative risk of disagreeing or being unsure about fertility desires compared with couples who both wanted a child. Across all three domains, the pattern was the same: more stress did not simply tip couples toward wanting fewer children; it made them less able to converge on any single answer at all.

The gender-stratified individual models revealed a subtler and more surprising asymmetry. For men, their own economic, health, and relationship stress showed no significant association with their fertility desires. The one partner effect that mattered was relational: men whose female partner reported relationship stress at least one standard deviation above their own expressed significantly lower desire for a(nother) child than men whose partners’ stress was similar to theirs. For women, their own stress in any domain was likewise unrelated to their desires, but the partner effect ran in an unexpected direction. Women whose male partners reported lower health stress than they did had lower fertility desires, contrary to the hypothesis that a more-stressed partner would dampen desire. The authors offer two plausible interpretations. A man who is comparatively unworried about pandemic health threats may be perceived as someone who will not take health precautions seriously, raising doubts about whether he would attend to children’s health needs; alternatively, women may have been especially conscious of pandemic-era policies that excluded partners from prenatal visits and delivery rooms, coloring how they weighed their partner’s stance.

Consistent covariates ran through the models. As the younger partner’s age rose, couples were more likely to agree they did not want a child, and married couples were less likely to want a child together than cohabiting couples. Couples with one child were more likely to want another than childless couples, while experienced financial hardship in the past month increased the odds of discordance or uncertainty rather than pushing couples cleanly in either direction. Among men, financial hardship was actually associated with higher fertility desire, while for women, part-time employment relative to full-time work significantly raised desire.

The authors are candid about limitations. The data are cross-sectional, so causality cannot be established, and the study cannot follow couples to see whether a child was eventually born. The dyadic subsample may overrepresent stronger relationships, as suggested by the low average relationship stress, and the pandemic-specific framing may limit generalizability, though the researchers argue that COVID-19 merely spotlighted a broader and arguably permanent feature of contemporary life: pervasive, multidimensional uncertainty. They also note that the U.S. context, with no universal health care and weak parental supports, may make fertility decisions especially sensitive to stress, and that future comparative work should test whether stronger social safety nets buffer these effects. The study excluded same-gender couples, for whom pathways to parenthood involve additional external gatekeepers such as fertility clinics and surrogates that the pandemic itself disrupted.

The implications reach well beyond the pandemic window. The findings vindicate the Narratives of the Future framework in a non-European setting and demonstrate that subjective well-being concerns operate across health and relational domains, not merely economic ones. They also carry a methodological lesson: proxy reports of a partner’s characteristics are demonstrably imprecise, and the gendered partner effects documented here, visible only because both partners were surveyed, would have been invisible in conventional single-respondent studies. Most fundamentally, the research reframes low fertility not as a simple calculus of income and employment but as a negotiation between two people each processing stress through their own lens, whose disagreement itself may be one of the quiet engines of America’s stalled birth rates. As the authors conclude, in an era when uncertainty from climate anxiety to threats to reproductive autonomy increasingly weighs on would-be parents, understanding how couples jointly make sense of an unpredictable future has never been more urgent.