The Career Return Arrives in Waves
The U.S. evidence on women’s education becomes most revealing when it is read as a sequence of ages rather than as one career average. The question is not simply whether women just before the California school-entry cutoff earn more; it is whether the education-induced advantage appears early, accumulates into mid-career, and remains intact when a first birth changes the demands around work. Reading the estimates this way gives the reader a more realistic judgment about what a return to education looks like in a life.
From a modest start to a mid-career gain
The women on the more education-favorable side of the cutoff have better labor-market outcomes across the observed ages, but the size and location of the difference change. Across ages 16 to 40, the study estimates an annual earnings gap of about $748 and an employment gap of about 0.7 percentage points. Those are local discontinuities: they compare women close to the birthdate boundary, not all college attendees with all non-attendees. Their importance comes from the way they are distributed over time.
In early career, women born just before the cutoff earn about $601 more, or roughly 6.9 percent relative to the local comparison group. By mid-career, the dollar gain is about $783, or 3.8 percent because the baseline level of earnings is higher. The percentage does not rise even though the dollar amount does; both the numerator and the scale of the comparison matter. The estimate is consistent with a career that begins with an education-related advantage and then grows in absolute value as earnings trajectories diverge.
Employment follows a related but not identical path. Around age 20, the estimated employment effect is close to zero. It rises to roughly 0.5 to 0.8 percentage points by the early thirties, producing a mid-career effect of about 0.7 points. This is the extensive margin: the likelihood of working. It should not be confused with the intensive margin, such as hours worked by people who are already employed. The study also finds higher earnings conditional on working, so the result is not explained only by bringing a few additional women into employment while leaving the pay of workers unchanged.
That distinction matters for interpretation. If education merely increased labor-force participation, the average earnings gain could arise from a composition change: more women would have positive earnings, but their jobs might not pay more than before. The conditional-earnings estimates point to a second component. Among women who work, the occupation and pay associated with the cutoff also improve. At the same time, the administrative data do not measure every aspect of labor supply equally well, so the evidence should be described as higher earnings both overall and among workers, not as a complete account of hours, effort, or satisfaction.
The location of the earnings gains changes as well. Early effects cluster near the poverty threshold, around $25,000 in annual income. By mid-career, gains center closer to $40,000, approximately the median income for women in the comparison used by the study, and spread upward toward higher incomes. The probability of earning above the poverty line increases by about 1.4 percentage points in mid-career compared with about 0.7 points earlier. In practical terms, the return is not just a larger paycheck for a stable group of already secure workers. For some women, it is a shift toward crossing a threshold of financial stability.
The changing denominator is part of the story. A $600 annual difference is large relative to earnings in the early twenties but represents a smaller percentage of a higher mid-career baseline. That does not make the later estimate weaker; it means that the same kind of comparison has a different economic meaning at different ages. A woman near the beginning of working life may experience the early effect as help with entering or staying in work, while the mid-career effect may matter more for household resilience, debt, housing, or the ability to absorb the costs associated with children. The study cannot observe every use of additional income, but its distributional estimates show that the threshold context changes over time.
Household earnings move in much the same direction as individual earnings. The study reports household-income effects that are close to the individual effect, including gains of roughly $800 to $1,100 in the early thirties. That similarity suggests little change in spousal earnings around the cutoff. The increase in household resources therefore appears to come primarily through the woman’s own labor-market position, rather than through a change in the economic status of her partner. This is important for the intergenerational question: additional resources available around a child’s early years are connected to the mother’s career, not merely to a different household pairing.
The age profile also helps distinguish a schooling pathway from a purely mechanical advantage of starting work sooner. Earlier school entry can allow earlier labor-market entry and therefore more initial experience. Such an effect should be strongest at the beginning and tend to fade as the experience gap narrows. The observed pattern runs in the opposite direction: the early gain is present but smaller in dollars, and the mid-career gain is larger. The credential evidence discussed previously makes education the authors’ leading explanation, while the life-cycle shape is consistent with education opening paths whose value emerges as women progress into more stable or higher-status roles. The timing supports that interpretation, but does not by itself prove every step between schooling and earnings.
This life-cycle map is not a promise that every stage improves for every woman. It records the pattern of the discontinuity: an early advantage, a stronger mid-career difference, and no detected collapse at the first birth. The separate “motherhood penalty” state is essential. A woman can face a substantial earnings or employment cost associated with having a child and still retain an education-induced advantage relative to another woman who faced the same transition.
Childbirth tests whether the gain survives
Motherhood is a demanding test of career durability because the arrival of a first child can change work hours, job choice, employment continuity, and the resources needed to remain attached to the labor market. If the education-induced gain placed women in jobs that were especially difficult to combine with children, the difference might shrink after childbirth. If it raised earnings enough to make childcare or other supports more affordable, the difference might grow. It might also remain unchanged if the cutoff shifted the level of the career path but not women’s response to motherhood.
The study compares outcomes around a first childbirth to distinguish these possibilities. In a balanced panel, women born just before the cutoff earn approximately $750 more in each of the three years before the birth and the four years after it. Employment displays a similarly flat pattern. There is no statistically significant break in the education-related advantage after childbirth. Estimates using the full sample of births are similar and somewhat more precise. The result is therefore persistence through a transition, not a claim that the transition itself is harmless.
This distinction between level and penalty is easy to lose. The cutoff does not appear to change the motherhood penalty—the drop in work or earnings associated with having a child—because the earnings difference between the two cutoff groups does not change sharply after the birth. Instead, the group with the additional education begins the transition from a higher career position and remains about that much ahead. More-educated mothers consequently have greater financial resources during pregnancy and the early years of a child’s life, even though motherhood can still impose a cost on both groups.
The fertility checks make that comparison more credible. Around the cutoff, the study finds no significant change in the likelihood of having any birth, a first birth, or a second birth, and no significant change in the average age at those births in the baseline samples. If the rule had caused one group to select into motherhood much more often or at systematically different ages, the apparent post-birth career difference could have reflected a different mix of mothers. The balance evidence does not eliminate every selection concern, but it reduces the most direct version of that explanation.
The balanced childbirth analysis also changes what a flat profile means. It is not simply comparing younger mothers with older mothers at different points in their lives; it follows the relevant mothers across a common window around the birth. That makes a sudden divergence after childbirth easier to detect. Its absence is consistent with the education-related advantage surviving the transition, but it cannot reveal whether the reason is childcare affordability, job flexibility, occupational selection, or a combination of factors. A stable gap is evidence of persistence, not a complete account of resilience.
The data also show why the conclusion should be stated carefully. The analysis follows women who are observed within a California birth-records and administrative-earnings framework, with sample restrictions designed to capture likely exposure to California’s school-entry rules. The childbirth comparison is about mothers in that setting. It does not establish that education eliminates career interruptions for mothers generally, that every form of college makes work more compatible with children, or that an increase in income automatically produces better child outcomes. Those are separate questions requiring different evidence.
What the finding does establish is more specific and more useful. The local education shift is associated with a labor-market gain that remains visible during pregnancy and after a first birth. The gain is not confined to a short window before women have children, and it is not erased by the transition that often widens gender differences in work and pay. This persistence turns the mother’s career into a possible bridge between education and children’s early environment: a career that remains stronger around childbirth can bring more resources into the period when prenatal and infant investments matter.
The bridge should not be treated as a single causal pipe. Earnings can affect housing, food, transport, medical choices, and the ability to purchase time or care. Job status and working conditions can affect stress, social support, and access to information. But the U.S. study measures these channels with different degrees of directness. It observes earnings and employment in administrative records; it assigns many job characteristics from occupation-level data; and it observes smoking, program participation, prenatal visits, and infant birth outcomes in birth certificates. The fact that the outcomes line up in time is evidence for a connected life-cycle pattern, not proof that one unmeasured feature of a job caused a particular infant result.
The growing return also changes how one should evaluate education policy or personal educational choices. An immediate comparison of tuition and starting wages may understate the payoff if education influences the occupation a woman reaches several years later and if that position supports continued employment after motherhood. On the other hand, a delayed gain can be difficult for a household to finance, and the relevant credential may be a two-year program or some college rather than a generic “college degree.” The study does not evaluate a complete policy package, so it cannot tell us which subsidy, childcare arrangement, or credential program would reproduce the result. It does show why those decisions should be judged against the full timing of work and family life.
The career evidence therefore has two parts. First, education is followed by higher earnings and somewhat greater employment, with the absolute gain strengthening into mid-career and reaching more women above a poverty threshold. Second, the gain survives the first birth rather than disappearing at the moment when careers are most vulnerable to interruption. That still leaves an unanswered question: what kind of work produces the durable difference? If the women are simply working more for more pay, the return is one thing. If they are also moving into work with higher status, meaning, better peers, or different demands, the value of education is larger and more complicated than the earnings curve alone suggests.