All topics › Parenting decisions family systems
How does sharing parental leave (including fathers taking leave) affect children, mothers, and fathers?
How does sharing parental leave — including fathers taking more of it — affect children, mothers, and fathers? Is there evidence that who takes the leave actually changes how a family works?
The best evidence we have comes from Nordic countries that reserved part of paid leave for fathers ("daddy quotas"): those policies genuinely increased how much leave fathers took, and the evidence suggests fathers who take leave stay more involved in caregiving afterwards [1][2][7]. But the downstream effects are mixed and contested: one Norwegian study found fathers' future earnings fell 2.1% (read as time shifting homeward) [1], while another study of the same reform found no earnings effect at all [4] — and evidence on mothers' careers and relationship stability genuinely conflicts [3][4][5]. The UK's Shared Parental Leave exists, but only about 2–5% of eligible fathers use it [9][10], so we have no good UK evidence on its effects.
Norway's 1993 reform is the most-studied event in this field. It reserved 4 of 42 paid leave weeks for the father — "use it or lose it" — and uptake rose by roughly 50 percentage points [1]. Because the reform switched on for babies born on or after 1 April 1993, researchers can compare families just either side of the date: a natural experiment.
Two strong studies of this same reform disagree on earnings. Rege and Solli (2013) found that the four weeks of paternity leave decreased fathers' future earnings by 2.1%, with time-use data showing treated fathers spent significantly more time with their children afterwards [1]. Their interpretation: fathers shifted effort from market to home production — the earnings "cost" is evidence of involvement, not harm. But Cools, Fiva and Kirkebøen (2015), using the same reform, found no negative effect on fathers' earnings and no effect on parents' labour supply [4]. A Norwegian statistics review attributes the conflict to different identification strategies and sensitivity to time trends [4]. This is the literature arguing with itself in public — take both findings seriously.
On the division of labour, Kotsadam and Finseraas (2011) found that 14–15 years after the 1993 reform, exposed parents reported a more equal division of household tasks [2] — though the study has a real weakness: child age correlates perfectly with treatment, so the effect could reflect time trends rather than the reform. In Québec, where a similar paternity-leave reform was introduced, Patnaik (2019, published) found the reform substantially raised fathers' participation — driven by higher benefits and the "daddy only" label — and presented causal evidence that paternity leave reduces sex specialisation in home and market production long after the leave period [8].
Sweden's 1995 "daddy month" evaluation is the cautionary tale: a strong difference-in-differences design found no behavioural effects — no change in leave to care for sick children, no effect on either parent's long-term wages or employment [3]. And a 2009 Norwegian expansion of the quota increased fathers' uptake but had no lasting labour-market effects, with a small short-run dip in mothers' earnings [5]. The pattern across reforms: reserving leave for fathers reliably increases take-up; everything after that is context-dependent.
Across Sweden, the UK, and the US, survey studies find that fathers who take longer leave are more involved in childcare months and years later [6][7]. In the US Fragile Families study, ~90% of resident fathers took some leave (mostly ≤2 weeks), and those taking 2+ weeks were more involved at 9 months [6]. A 2012 OECD review synthesised these findings across four countries — but noted the Australian exception, where longer leave showed no such association [7].
The selection problem here is obvious: fathers who take long leave tend to be the ones already inclined toward caregiving, with flexible jobs and egalitarian attitudes. The quasi-experimental studies (Norway, Québec) are the reason we can say leave-taking itself probably moves the needle — but even there, the effects are modest and uneven.
Nobody has credibly shown that fathers taking leave directly improves child outcomes. The chain is: leave → more involved fathers → father involvement is linked to positive child outcomes in longitudinal studies (e.g. Sarkadi et al.'s 2008 systematic review) [7]. Each arrow in that chain is weaker than you'd like. Be honest about this: "fathers' leave is good for children" is a reasonable inference, not a measured fact.
This is where families most want answers, and where the literature most frustrates:
The UK's Shared Parental Leave (introduced 2015) lets parents share up to 50 weeks of leave and 37 weeks of pay; statutory Shared Parental Pay is £194.32/week or 90% of earnings, whichever is lower (2026–27) [11]. Each parent can book up to 3 blocks, with 8 weeks' notice; the mother must end maternity leave (after the mandatory 2 weeks) to create the shared pot [11].
Uptake is tiny: official surveys found 1% of mothers and 4% of fathers took SPL [9]; an ISER briefing puts eligible-father use at ~2–5% [10]. The reasons show up clearly in the data: 35% of fathers who didn't take statutory leave said they couldn't afford to (vs 11% of mothers), and in ~2 in 5 cases, not being able to afford to stay on leave was the reason for returning [9]. Statutory paternity pay was worth about 46% of the National Living Wage for a 35-hour week in the JRF analysis (£184.03/week at the time) [13] — the UK has among the least generous statutory paternity pay in the OECD [10]. And a Bath/Cardiff evaluation found SPL barely changed total uptake or leave length compared with the pre-2015 regime [10].
Bottom line: we have no credible UK evidence on what SPL does to families, because almost nobody uses it.
This is a money decision first, a values decision second. The research is blunt: fathers don't take leave mostly because they can't afford to — 35% said so directly [9]. At £194.32 a week, a father on an average salary faces a steep pay cut for each week of SPL, and fathers are typically the higher earner with less generous employer top-ups [10]. Before philosophising about involved fatherhood, do the household spreadsheet. The Nordic evidence says "use it or lose it" quotas work because they're paid well enough to use.
Employer top-ups are the real policy. The statutory floor is nearly irrelevant for many families; what matters is whether either parent's employer enhances pay. Ask both employers, in writing, before the baby arrives — and check for strings (some top-ups must be repaid if you don't return).
Involvement begets involvement. The most consistent finding in the whole literature is that fathers who take leave stay more hands-on afterwards [1][2][6][7]. If you want a household where caregiving is genuinely shared, the leave decision is the earliest lever you have — and the Nordic evidence suggests it isn't just selection.
The earnings "cost" is ambiguous. A 2.1% dip in fathers' future earnings [1] sounds alarming until you read the interpretation: it reflects time moved from work to home. Whether that's a cost or a benefit depends entirely on what your family values — which is precisely why the conflicting Norwegian study [4] matters: even the direction of the effect is uncertain.
For mothers, the honest answer is "we don't know yet." The theory — fathers take more leave, mothers return to work earlier and earn more — is plausible and supported in some countries [8], contradicted in others [3][4]. Don't plan your family's finances around a career boost the evidence can't promise.
The default is powerful. Where leave is individual and transferable (as in the UK), mothers take it; where it's reserved for fathers, fathers take it [1][10]. If you want to share, you have to actively opt in against the grain of the system — the low UK take-up isn't a verdict on the idea, it's a verdict on the design.
This topic has no binary child-health outcomes with base rates — the outcomes are behavioural and economic. The honest presentation is descriptive:
| Finding | Absolute terms | Evidence rating |
|---|---|---|
| Norway 1993 quota → fathers' leave uptake | Rose by ~50 percentage points [1] | B (quasi-experimental) |
| Fathers' future earnings (Rege & Solli) | −2.1% [1] | C (contested by [4]) |
| UK SPL take-up among eligible fathers | ~2–5 in 100 [9][10] | B (official statistics) |
| Fathers citing affordability for not taking leave | 35 in 100 fathers vs 11 in 100 mothers [9] | B |
| UK fathers taking ≥2 weeks with more involvement at 9 months (US study) | Association only; ~90% took some leave, mostly ≤2 weeks [6] | C (observational) |
| Statutory paternity pay vs National Living Wage (35h week) | ~46% [13] | B (official figures) |
This topic is mostly a household and financial decision, not a medical one — but: