Divorce optimism widens the gender earnings gap
Underestimating divorce risk encourages couples to specialize more and have children earlier, a Norges Bank working paper finds. Secondary earners accumulate less human capital, leaving them financially less prepared if the marriage ends.
Marital confidence carries career costs
Ursula Berresheim, David Koll and Mariia Kovaleva combine a survey of 2,595 married Americans, conducted in 2024–2025, with a household life-cycle model.
Respondents put their ten-year divorce risk 16 percentage points below comparable couples; two-thirds leave their beliefs unchanged after receiving population divorce statistics.
Against a statistical benchmark, 74 percent underestimate their lifetime risk, by 7.8 percentage points on average and 15.4 points among those married under four years.
In the model, correcting beliefs reduces the working-life gender earnings gap by about 6 percent and the human-capital gap by about 7 percent.
It reduces the child earnings penalty by about 3 percent in years zero through five after birth and 7 percent in years six through ten.
Early choices outlast the bias
Modeled optimistic couples have their first child at age 27.17 rather than 28.27 and have about 8 percent more children.
Their belief bias has a 2.6-year half-life, but secondary earners' lost human capital persists after divorce.
The average lifetime welfare cost equals 0.27 percent of consumption, rising to 8.63 percent for bad matches; good matches yield a 4.60 percent gain.
Additional children offset much of the average loss.
Divorced optimistic wives consume about 3 percent less than realistic counterparts.
The benchmark has limits
The authors note that survey correlations do not establish causality.
Their statistical benchmark uses older marriage cohorts and may overstate younger respondents' divorce risk.
The model assumes shared spousal beliefs and equal asset division, while excluding remarriage, alimony and child support.