The National Survey of Family Growth: Unpacking the Impact of Age at Marriage and Premarital Children on Divorce Risk

Study: National Survey of Family Growth (NSFG) 2015–2019 | Conducted by: Centers for Disease Control and Prevention (CDC)

Study Description

The National Survey of Family Growth (NSFG), conducted by the Centers for Disease Control and Prevention (CDC), is one of the most comprehensive sources of data on family dynamics in the United States. Running since 1973, the NSFG provides insights into how factors like age at marriage and premarital children influence marital stability. The 2015–2019 cycle, in particular, offers updated findings from thousands of couples, showing that marrying young or having a child before marriage significantly raises divorce rates.

Research Findings

The data reveals that couples who marry under 23 have a 10-year divorce rate of 40–50%, compared to the U.S. average of 28%. That's a relative risk (RR) of 1.8–2.5, meaning 80–150% higher odds of splitting. Younger couples often face challenges like emotional immaturity, financial instability, and rushed commitments. On the flip side, marrying at 25 or older drops the rate to 20–25% (RR 0.6–0.8, a 20–40% reduction).

The study also highlights premarital children, often leading to "shotgun marriages," where the 10-year divorce rate climbs to 40–50%. These unions struggle with added stress from parenting without a solid foundation. Without a premarital child, rates drop to 22–25% (RR 0.8–0.9, –10–20%).

Experimental Setup

The NSFG is a national probability survey, selecting participants to represent the U.S. population. The 2015–2019 cycle interviewed over 13,000 adults aged 15–49 from diverse demographics, using in-person and computer-assisted methods for accuracy. Respondents provide detailed timelines of marriages, divorces, pregnancies, and other events. Researchers use Cox proportional hazards models to analyze divorce timing over 10 years, controlling for confounders like socioeconomic status, race, education, and religion. This setup isolates the effects of age and premarital children, tracking how they predict marital dissolution while adjusting for overlaps.

Drawbacks/Limitations on Finding

While robust, the NSFG has limitations. Self-reported data can have recall bias (people might misremember dates or downplay divorces). The sample focuses on younger adults (15–49), so findings may not generalize to older marriages. Cultural shifts since 2019 (e.g., more cohabitation, delayed marriage due to COVID) aren't captured, potentially underestimating effects in Gen Z. Prevalence adjustment assumes average exposure, but regional differences (e.g., higher teen marriages in the South) aren't modeled. Finally, causality isn't proven—young age might correlate with unmeasured factors like impulsivity. Despite this, the study's national scale and controls make it highly reliable for population estimates.

Calculator Integration

At Odds on Life, we use NSFG data for age and premarital child factors. Age logit is +0.916 for <23 (high-risk) and -0.405 for ≥25 (protective), with +0.182 extra if the wife is younger. For premarital child, logit +0.588 (yes) and -0.182 (no). These are standalone factors with no cap, as NSFG shows low overlap with others like income (already controlled). This ensures accurate, non-inflated estimates—e.g., a young couple with a premarital child might see 45–55% risk, matching NSFG data, rather than 70%+.

Study References

Related Factors

This study directly informs the calculator's assessment of:

These factors work independently in our calculator, as the NSFG data shows low overlap with other demographic factors that are already controlled for in the study's methodology.

← Back to Reference Go to Calculator