All topics › Parenting decisions family systems
Does the timing of returning to work after a baby affect child outcomes or parental wellbeing?
Does the timing of returning to work after a baby affect child outcomes — or your own wellbeing as a parent? And is there a "right" month to go back?
There is no proven "right" time to return, and no study can prove one — you can't randomise when parents go back to work. The best long-term study (the US NICHD study of early childcare) finds that full-time work in the first year is associated with slightly lower cognitive scores and a little more caregiver-reported behaviour problems, but the differences are small, heavily tangled up with family income and childcare quality, and the study's own modelling finds the overall net effect is roughly neutral [1][3]. What the evidence does show clearly: returning earlier and full-time is linked to shorter breastfeeding [4], and longer leave is linked to fewer maternal depressive symptoms [5][6]. The honest bottom line is trade-offs, not a verdict.
The landmark study here is the NICHD Study of Early Child Care and Youth Development, re-analysed in a 2010 monograph by Brooks-Gunn, Han and Waldfogel [1]. It followed children whose mothers worked full-time (55% of the sample), part-time (23%), or not at all (22%) during the child's first year, with an unusually rich set of controls for family background.
For non-Hispanic white children, full-time (but not part-time) first-year employment was associated with lower scores on some cognitive measures at ages 3, 4½ and first grade — effect sizes of −0.10 to −0.18 standard deviations [1]. To put that in plain terms: a shift of about a tenth to a fifth of a standard deviation is the kind of difference that is statistically detectable in a large study but barely visible in an individual child. There were no significant associations for African-American children (a much smaller sample), and no link between first-year employment and attachment security [1].
On behaviour: full-time work begun by 3 months was associated with more caregiver- and teacher-reported externalising problems at 4½ and first grade [1]. In absolute terms, the share of children above the clinical cut-off for behaviour problems (maternal report) was 2% for full-time vs 1% for part-time at age 3, 4% vs 2% at age 4½, and 3% vs 1% at first grade [1]. So we are talking about small numbers of children and small differences — not a cliff edge.
Crucially, the authors' structural-equation modelling found the overall net effect of first-year maternal employment on cognitive outcomes was neutral: negative direct associations were offset by positive indirect ones — working mothers had higher earnings, used more centre-based care, and part-time working mothers were rated as more sensitive [1]. And the whole literature comes with the same giant caveat: mothers who work full-time in the first year are a highly selected group — on average more educated, higher-earning, and using different childcare than mothers who don't [1][3]. No statistical control can fully unpick that.
Ruhm's (2004) analysis of the US National Longitudinal Survey of Youth tells a similar story: 20 extra hours a week of maternal work during the first three years was associated with test-score reductions of 0.04–0.11 standard deviations at ages 3–6 — small effects, concentrated among children whose mothers worked long hours, with evidence of reverse causation (mothers tended to work less after a child scored poorly) [3].
The UK Millennium Cohort Study gives the cleanest numbers on this [4]. Among 6,917 employed mothers, the share still giving any breast milk at 4 months was:
So mothers who went back by 4 months were roughly half as likely to still be breastfeeding at 4 months as those who went back at 7–8 months or later. Part-time and self-employed mothers breastfed longer than full-time mothers (41 in 100 for self-employed vs 25 in 100 for full-time), and mothers who said they returned for financial reasons breastfed less (25 vs 32 in 100) [4]. These relationships survived adjustment for education and socio-economic status — but residual confounding is still possible, and the direction can run both ways (a mother planning to stop breastfeeding may choose to return earlier).
Chatterji and Markowitz, using US data with an instrumental-variables design, found that longer maternity leave — paid or unpaid — was associated with fewer depressive symptoms, a lower likelihood of severe depression, and better self-reported overall health [5]. But their earlier work adds an important nuance: later return was linked to fewer depressive symptoms but not to a lower probability of being a likely case of clinical depression [6]. A 2023 systematic review found the wider literature mixed: only 6 of 17 studies supported a link between longer leave and fewer depressive symptoms, though two US studies found returning after 12 weeks (vs before) was linked to a lower likelihood of major depression [7].
Meanwhile, in the NICHD study, mothers who worked full-time in the first year reported more depressive symptoms than non-working mothers at every assessment (effect sizes 0.21–0.36), while part-time mothers were barely elevated (0.07–0.09) [2]. But which way the arrow points is genuinely unclear: pre-existing differences in mental health shape who returns early and full-time, and the study can't settle that.
UK employees are entitled to up to 52 weeks of maternity leave, with Statutory Maternity Pay for up to 39 weeks: 90% of average weekly earnings for the first 6 weeks, then £194.32 a week or 90% of earnings, whichever is lower (rate for 5 April 2026–3 April 2027) [8]. The last 13 weeks of leave are unpaid unless your employer tops up. You can also work up to 10 paid "keeping in touch" (KIT) days during maternity leave without ending it [9].
This decision lands on parents' bodies, minds, and bank accounts — and the evidence is more useful here than on child outcomes.
Money is the unavoidable frame. In the NICHD study, full-time first-year working mothers were substantially more likely to be in the top third of earners later on (effect sizes around 0.35–0.39) — employment compounds financially [2]. In the UK, the pay structure does the talking: after 6 weeks, statutory pay drops to £194.32 a week, and the final 13 weeks are unpaid [8]. In the Millennium Cohort, mothers who returned for financial reasons breastfed for less time — a hint that constrained choices, not preferences, drive a lot of early returns [4]. If you can afford to stay longer, that's a financial privilege, not a parenting achievement; if you can't, that's economics, not a failing.
Mental health is genuinely two-sided. Longer leave is associated with fewer depressive symptoms [5][6], but employment itself isn't the villain: part-time return looks neutral-to-positive in the NICHD data [2], and many parents report that work restores identity, adult contact, and self-esteem after the intensity of early months. The honest evidence position: we have decent data that very early return is associated with more depressive symptoms, and thin, mostly cross-sectional data on the identity benefits of work. Both can be true, and which dominates is personal.
Sleep and load barely feature in the research. No study in this topic measured parental sleep as an outcome of return timing — a striking gap given that the return-to-work months coincide with still-fragmented infant sleep. Treat anyone who claims a definitive answer on "work vs sleep" with suspicion.
Relationships take the strain either way. The evidence on couple conflict and division of labour sits mostly in the parental-leave-sharing literature (see Dividing parental leave), but one relevant finding: mothers whose partners took no leave at all after the birth reported more depressive symptoms [5] — the return decision is a household decision, not an individual one.
Fathers and partners are nearly invisible in this literature. Almost every study here is about mothers' employment and mothers' outcomes. If you're the non-birthing parent weighing when to return, the evidence base has very little to say to you directly — that's a gap in the science, not a finding.
| Decision | Outcome | Effect in absolute terms | Evidence rating |
|---|---|---|---|
| Returning ≤4 months vs 7–8+ months | Still breastfeeding at 4 months | ~16–18 in 100 vs 39 in 100 (UK cohort, adjusted) [4] | B |
| Full-time vs part-time first year | Child above clinical behaviour-problem cut-off (maternal report) | Age 3: 2 vs 1 in 100; age 4½: 4 vs 2 in 100; first grade: 3 vs 1 in 100 [1] | C |
| Full-time first-year employment | Cognitive test scores | ~0.10–0.18 SD lower on some measures (white children; small, confounded) [1]; 0.04–0.11 SD per 20 extra hrs/week [3] | C |
| Longer vs shorter maternity leave | Maternal depressive symptoms | Fewer symptoms with longer leave [5]; but no clear effect on likelihood of clinical depression [6] | C |
| Full-time first-year employment | Maternal earnings later | Substantially higher — more likely in top third of earners [2] | C (observational) |
| Returning for financial reasons | Breastfeeding ≥4 months | 25 vs 32 in 100 [4] | C |