Shared Custody, Sole Custody, and Everything In Between: What the Evidence Supports

Few divorce questions feel as consequential as custody, and few are surrounded by as much unreliable commentary. The research literature on custody arrangements is substantial — enough to answer some questions confidently, and enough to confirm that other questions remain genuinely open. This article walks through both: where the evidence converges and where it still doesn't.

The language, briefly

Custody terminology varies by jurisdiction and is often confusing. In the research literature, three distinctions matter.

• Legal custody (decision-making authority) versus physical custody (where the child lives). These are often split differently.
• Sole physical custody (child lives with one parent, visits the other) versus shared physical custody (child lives meaningfully with both). Research operationalizes 'shared' variously — commonly 35%/65% or more time with each parent, though 50%/50% is the modal arrangement in shared-custody samples.
• Joint legal custody (both parents have decision-making authority) is separate from physical arrangement. In many jurisdictions, joint legal is the default even when physical is primarily with one parent.

What the shared-custody research shows

The shared physical custody literature has grown substantially since 2010. The largest single contribution is the Swedish research program (Bergström and colleagues) using population-level data on children's wellbeing across custody arrangements. The consistent finding: children in shared physical custody report better psychological outcomes than children in sole custody, and comparable outcomes to children in intact families.

This finding held across multiple methodologies (cross-sectional and longitudinal), across multiple countries (Sweden, U.S., Netherlands, Australia), and after controlling for obvious confounders (family income, parent education, parental conflict at separation). Selection effects — the worry that the parents who choose shared custody are systematically different from those who don't — remain a methodological concern but have been addressed in several analyses without overturning the main finding.

The conflict exception

Shared custody's favorable outcomes reverse when exposure to parental conflict is high. The pattern across studies is clear: shared custody in low-conflict contexts is associated with good child outcomes; shared custody in high-conflict contexts can produce worse outcomes than sole custody with restricted contact, because more time together means more conflict incidents for the child to witness.

This interaction is why the academic consensus — to the extent one exists — is closer to 'shared custody is associated with better outcomes for most children, provided conflict can be contained' than 'shared custody is always best.' The containment question matters enormously. A well-structured parallel-parenting arrangement can contain conflict within a shared-custody schedule. A contested, conflictual shared-custody arrangement often cannot.

The overnights debate

The most visible live debate in custody research concerns overnights for infants and toddlers. Two bodies of research reach different conclusions, both based on real evidence, and the disagreement has practical consequences for parenting plans.

The McIntosh position

Australian research led by Jennifer McIntosh in the early 2010s reported that infants and very young children who had overnight time with a non-primary parent showed elevated stress markers and disrupted attachment patterns compared to age-matched peers without overnights. The implication for policy and parenting plans was that overnights for very young children should be deferred until attachment security was established.

The Warshak consensus

A 2014 consensus report organized by Richard Warshak, endorsed by 110 researchers across the field, argued that the McIntosh findings did not support the policy conclusions being drawn from them — that sample sizes were small, methodologies limited, and the broader attachment and custody literature did not support restricting overnights for infants in otherwise healthy co-parenting relationships. The consensus position supported overnights from very early ages when both parents were available caregivers.

What to make of this as a parent

This debate has not been resolved, and parents making overnight decisions for infants are making them under genuine empirical uncertainty. What does seem well-supported, independent of the specific schedule:

• Young children benefit from frequent contact with both parents when both are available caregivers. This is not contested.
• Consistency and predictability of caregiving matter enormously at this age, and frequent schedule changes can be destabilizing.
• The quality of caregiving matters more than the quantity at any specific location. A responsive, attuned parent across three daytime visits per week may be more protective than an under-resourced overnight arrangement.
• As children move past infancy into toddlerhood and early childhood, the evidence more clearly supports shared overnight arrangements for most families.

What courts actually weigh

Every jurisdiction applies some version of the 'best interests of the child' standard, but what that translates into operationally varies more than parents realize. Understanding your jurisdiction's specific factors is usually more useful than arguing from the general research.

Common factors across jurisdictions

• The child's existing relationship with each parent and the history of caregiving responsibilities.
• Each parent's capacity and willingness to support the child's relationship with the other parent. (This 'friendly parent' criterion is explicit in many statutes.)
• Stability of each household — housing, employment, routines, schools.
• The child's adjustment to home, school, and community.
• Physical and mental health of all parties.
• Evidence of family violence or abuse — which typically dominates the analysis when present.
• The child's preferences, weighted by age and maturity.
• Geographic proximity of the two parents and practical feasibility of schedules.

Jurisdictional variation worth knowing about

Several U.S. states have adopted presumptions in favor of shared parenting (Kentucky, Arizona among the strongest). Canada's Divorce Act amendments in 2021 replaced 'custody' and 'access' language with 'decision-making responsibility' and 'parenting time' and emphasized the best-interests standard with specific statutory factors. Australia's Family Law Act underwent major changes in 2024 removing earlier presumptions about equal shared parental responsibility. These shifts matter because the starting point affects the outcome, and the starting point varies.

Practical considerations that matter more than schedules

Most custody disagreements focus on the percentage split — 50/50 versus 60/40 versus every-other-weekend. The research suggests the more consequential variables are usually structural.

• Proximity. Schedules that require the child to travel far between households on school days are harder to sustain than schedules with short transitions. Proximity is often the hidden determinant of whether a schedule is actually workable.
• School stability. Enrolling the child in a school that both parents can reach, and keeping them enrolled through transitions, is more protective than arguing about schedule percentages.
• Predictability. Children cope better with consistent, predictable schedules than with frequently renegotiated ones. Many parents under-weight this relative to flexibility.
• Handoff design. Where, how, and how often children transition between households is a high-leverage variable. Structured, calm handoffs matter more than the specific number of nights.
• Developmental adaptation. As children age, optimal schedules change. The research supports revisiting parenting plans at developmental transitions (school entry, adolescence) rather than assuming a single arrangement works across childhood.

What the evidence suggests for your decision

If you and the other parent can contain conflict: the research favors shared physical custody (broadly defined) for most children. The specific percentage matters less than parents think. What matters is that both parents remain meaningfully involved, transitions are structured, and the child's routines stay stable.

If you cannot contain conflict: the research does not favor forcing a shared arrangement that will expose your child to more conflict. Sole physical custody with structured access may produce better child outcomes than a 50/50 arrangement that generates constant disputes.

If you have very young children: the overnight question is genuinely unsettled. Frequent contact with both parents, caregiving continuity, and stable transitions are the points of agreement across the debate. Most parenting plans for infants and toddlers can be designed responsibly around those points without waiting for the field to resolve the rest.