- At the individual level, marriage can't be expected to deliver the same benefits if doesn't include fidelity; at the societal level, marriage can't be expected to deliver the same benefits if it is fragile. Tweet This
- Torche and Abufhele lean closer to the conclusion that the marriage premium is entirely socially constructed than their analysis supports. Tweet This
Are the health risks associated with obesity going to diminish because most of us are emerging from the pandemic heavier than they went in? Don't answer too quickly: First, consider that some of the health risks are essential (e.g., ordinary movement taxes the heart more when body mass is greater), while others are socially constructed (e.g., stress depends on how much fat shaming society supports, and the resources for self-love society provides).
Thinking carefully about essentialism and social construction with respect to obesity and health will equip you to appreciate Florencia Torche and Alejandra Abufhele's work published in the American Journal of Sociology last month. They build a case that health risks associated with being born to a single mother (aka the marriage premium for infant health) depend on how society treats single mothers. This is quite like contending that the health risks associated with obesity depend on how society treats overweight people. Note that the contention is agnostic on whether or not there is an essential component—it merely asserts that the size of the difference between groups includes a socially constructed piece.
Torche and Abufhele add context to a debate that too often (but not always) focuses on causation v. selection, i.e., marriage inherently delivers a health premium v. healthier people are more likely to marry (and the same individuals would be healthier even if they had not married). The authors carefully show that selection explains only a small part1 of why the marriage premium for infant health dwindled as nonmarital births increased from 34% of all births in 1990 to 73% in 2016 in Chile. They then encourage us to think about marriage as a characteristic of individuals and societies. They focus on the potential for increases in nonmarital childbearing to rectify societal mistreatment of unmarried mothers: less stigmatization of unmarried births with all its labeling, status loss, and discrimination would translate into less internalization of society's negative responses, and therefore healthier infants at birth.2
As women who believe that marriage is inherently good, we often find ourselves treated as if a belief that greater body mass increases heart attack risk is tantamount to denying that fat shaming can contribute to depression. We instead embrace the reality that health has socially constructed components that change with context.
For instance, another reason why the marriage premium for infant health may have deteriorated as nonmarital births shift from being a minority share to a majority share is that marriage itself could confer fewer benefits when it loses some of its ability to confer security in individual lives. The rise in nonmarital childbearing has been part of a broader "retreat from marriage" that includes later marriage, more divorce, and less remarriage. When the majority of the adult life course is typically spent unmarried, people tend to plan for independence rather than interdependence, and entering marriage becomes less likely to be accompanied by a sureness that it will last.
Many students here at The Catholic University of America where we work are children of divorce, and even those with married parents deeply understand that staying married has become less common in recent generations. If part of why married mothers have healthier babies is that they benefit from confidence that the fathers will remain intimately involved with them and their shared children, then this benefit from marriage could diminish over time. Put differently, we share Torche and Abufhele's belief that prevalence of unmarried motherhood conditions the size of the marriage premium, but we think the mechanisms likely include increased stress among married mothers as well as decreased stress among unmarried mothers.
Another reason the marriage premium for infant health may have deteriorated as nonmarital births shift from being a minority to a majority share is that marriage itself could confer fewer benefits when it loses some of its ability to confer security in individual lives.
To make our values statement clear, we believe that the intrinsic good of marriage depends upon both individual commitment to the institution and societal commitment to the institution. At the individual level, marriage can't be expected to deliver the same benefits if doesn't include fidelity; at the societal level, marriage can't be expected to deliver the same benefits if it is fragile.
Torche and Abufhele characterize marriage promotion advocates as assuming that marriage is worth promoting because it is essentially good for those who marry and for their children; they contrast that with the contextual approach they take. We are advocates of marriage who embrace the contextual approach, and we have no qualms about admitting that promoting it in individual lives will not necessarily deliver all its potential benefits.
We do, however, contend that Torche and Abufhele lean closer to the conclusion that the marriage premium is entirely socially constructed than their analysis supports. For instance, they speculate that the marriage premium for older children's outcomes will dwindle the way the marriage premium for infant health has. Their reasoning is that whereas newborns are affected by stigma via maternal stress during gestation, children of unmarried mothers experience further stigma directly as they engage with society. This stigma-focused reasoning neglects the fact that caregiving matters for post-birth outcomes in a way it cannot for at-birth outcomes.
Marriage inherently supports caregiving by helping keep fathers involved in children's lives. Of course, the extent to which it does so is socially variable, but we are emphasizing that some marriage premia have a larger essential component than others. Chilean data show that while nonmarital childbearing has been rising, the share of children living with both biological parents at age 12 has been falling. Cross-culturally, marriage is understood to be a greater commitment than cohabitation, and this commitment supports caregiving. The essential benefits of family stability across childhood cannot, of course, be measured in an analysis like Torche and Abufhele's that focuses on infant health at birth.
Further, one of their three analyses was likely biased by the sample selection involved when using individual woman fixed effects. While this statistical technique eliminates some sources of bias, it introduces others as explained here. Fixed effects could have been more profitably employed at the municipality level.3
In sum, we applaud Torche and Abufhele's focus on context demonstrating that the marriage premium for infant health has a socially constructed component. But we also want to emphasize how different this is from assaulting the claim that marriage is essentially good. They never make a claim tantamount to denying that greater body mass increases heart attack risk, but they do cast more doubt on the essential goodness of marriage than their analysis supports.
Laurie DeRose is a senior fellow at the Institute for Family Studies, Assistant Professor of Sociology at the Catholic University of America, and Director of Research for the World Family Map Project. Anna Barren is a graduate of the Philosophy program at Christendom College and is the administrator of the Sociology department at the Catholic University of America.
1. They rule out the possibility that the marriage premium dwindled because a greater share of nonmarital births occurred in cohabitation rather than to lone mothers by documenting that the proportion of nonmarital births in these two categories stayed the same while the overall share of nonmarital births skyrocketed. They also showed that a greater share of nonmarital births occurred to socioeconomically advantaged women over time, but the marriage premium would still have shrunken without this compositional shift.
2. They brilliantly argue against the possibility that something unrelated to normativity caused convergence of infant outcomes over time by showing that the health penalty associated with having no identified father on the birth certificate was not attenuated over the same period when health penalty for other nonmarital births shrunk considerably.
3. For Chile as a whole, they observed the marriage premium increasing as nonmarital births became more normative over time. When they compared normativity across municipalities, they restricted the time window instead of using observations from multiple time points that would have permitted statistically controlling for unchanging characteristics of geographic areas: if their finding had persisted with this control, it would have been far more convincing.