The Mixed Effects of Postponing Union Formation

Two articles on reproductive unions published this past spring have little in common except for their attention to a person’s age at first cohabiting union. Rolf Granholm, Anne Gauthier, and Gert Stulp analyzed data from the Netherlands to show that delays in union formation contribute to people ending up with fewer children than they intend to have; in another study, Opoku Adabor and Ankita Mishra analyzed data from Ghana to show that people delaying unions have healthier children. Of course, one research piece pointing to a negative outcome from delaying reproductive unions and another identifying a positive outcome from this delay do not need to be at odds any more than coffee causing jitters and helping with focus.

Country Context Matters

Before even considering the mixed outcomes, however, we need to consider whether a delay in the Netherlands means the same thing as a delay in Ghana. In both studies, the relationship behavior that was delayed was entering a union for the first time (marital or cohabiting). But the average age at first coresidential union is 23.7 years in the Netherlands and 19.6 years in Ghana. In Ghana, the most typical age for moving in with a partner is 18, and by age 20, most have formed their first union. This means that a woman entering first union when she is 21 does so early in the Netherlands and late in Ghana. These differences, of course, have a lot to do with economic development: it is relatively rare for Ghanaian women to pursue tertiary education (about a quarter do) and quite normative for Dutch women (94% do). The vast difference between Western Europe and West Africa is unsurprising given that unions usually commence after school completion. School is compulsory through age 18 in the Netherlands: if such a policy could be realistically implemented in Ghana, I am sure it would no longer be typical for unions to start at age 18.

Longer relationships, which are more likely to take place inside a stable marriage, are good for both fertility and reducing child mortality.

Age at First Union Makes A Difference

These absolute ages matter, as it turns out. Granholm et al.’s meticulous work in The Netherlands was based on simulations accounting for the fact that delaying union formation from 15 to 20 doesn’t have the same implications as delaying from 25 to 30. Implications for what? The “fertility gap”, i.e., the difference between intended and actual fertility. Dutch women, on average, have 0.34 fewer children than they intend to, or, more intuitively, about 1 in 3 women have one child fewer than they intend. This is a fairly typical fertility gap in Northern countries given that intended fertility still generally conforms to the 2-child norm while actual fertility falls short of replacement. Granholm et al. estimated that starting unions at age 15 would reduce the fertility gap 0.09 making it 0.25 children. In other words, if age at first union formation were decreased from 23.7 to 15, 1 in 4 people would have one child less than they intended instead of 1 in 3. Delaying average age at union entry beyond 23.7 matters more: increasing it to 28.7 years would add 0.23 children to the fertility gap, making it 0.58: a greater than 1 in 2 chance of falling short of the intended number of children. This age-dependent relationship makes sense: if the goal is to have two children, the first decade of union “delay” still results in unions occurring well before age-related conception chances start dropping (around age 32). Further delays are not as biologically innocuous.

Likewise, consider how absolute ages matter for understanding what it means for Ghanaians who delay unions to have healthier children. A sobering 13.5% of Ghanaian infants die before they are a month old; that is the average neonatal mortality rate in a population whose average entry age for first union is 19.6 years. I used Adabor and Mishra’s model to predict neonatal mortality if Ghanaian women formed first unions at the same average age as Dutch women (23.7 years): It was 6.9%—a dramatically lower rate still more than twice as high as in the Netherlands. 

The outcomes associated with union delays arise from the interplay between social context and biological realities.

However, the estimate if Ghanaian women were to enter unions at an average age of 28.8 years is equal to the current 3% neonatal mortality rate in the Netherlands. My calculations here result in unbelievable values because Adabor and Mishra’s model parameters were derived from data on the real experiences of a population that enters unions young and can’t be extrapolated like this. Which is my point. Because the authors of the Guana study calculated the actual neonatal mortality rate of those entering unions a year later than average (age 20.6) at about 11.5%, we can trust their conclusion—those who delay unions have healthier children—but we can’t predict mortality using values so far away from the actual data. 

What counts as a delay depends on social context. The outcomes associated with union delays arise from the interplay between social context and biological realities. We have no particular reason to believe that delaying unions by a year in the Netherlands (to an average age of 28.7) would improve infant survival at all, both because 1) survival chances are already excellent, and 2) they vary more at extreme reproductive ages (10-19 and 35+) than at more typical birthing ages. Nonetheless, that delay would decrease the chance that a Dutch women would bear all the children she intended to. Similarly, Ghanaians entering unions at 20.6 instead of 19.6 would be unlikely to increase the share not achieving their fertility intention while it would increase the share of surviving children. 

Stable Unions Matter Most

I close by highlighting a common finding from these two dissimilar studies: both found that stable unions are good for individuals. In the Netherlands, it wasn’t just that entering into a relationship later in life reduced the chance of achieving intended fertility—unions that failed to last also contributed to the fertility gap. In Ghana, the duration of unions before childbirth exerted an even more pronounced effect on neonatal mortality than the age at which women first cohabit. As the study notes: 

longer periods of relationship allow individuals to know each other, behave well, be mature and pool resources to support each other, thus improving the quality of the relationship and fostering stability. 

Longer unions, which are more likely to take place inside a stable marriage, are good for both fertility and reducing child mortality.

Laurie DeRose is a Senior Fellow at the Institute for Family Studies and Director of Research for the World Family Map Project.

*Photo credit: Shutterstock