If a parent in the middle of a separation reads only one body of research, it should be the research on interparental conflict. Across four decades, dozens of independent labs, meta-analyses of hundreds of studies, and longitudinal samples followed for thirty years, one finding keeps arriving in the same shape. It is the conflict between parents, not the separation itself, that does most of the damage to children.
This is not a soft-edged piece of clinical wisdom. It is one of the more replicated findings in developmental psychology. Knowing it changes what parents owe to their children during a separation, and it changes the benchmark for whether a coparenting relationship is "good enough."
The foundational finding
The first definitive meta-analytic statement came from Paul Amato and Bruce Keith in 1991. They pooled 92 studies comparing children of divorced and continuously married parents. Children of divorce scored lower across a range of outcomes, but the median effect was modest, and the most consistent support, by far, was for a family conflict explanation rather than for parental absence or economic loss (Psychological Bulletin, 110(1)).
Amato updated the analysis in 2001 with 67 more studies from the 1990s (Journal of Family Psychology, 15(3)). The update sharpened the point. Where marital conflict was overt, intense, chronic and unresolved, children appeared to be better off in the long run if the marriage ended than if the parents remained together. Where conflict was low, children appeared to be worse off following a divorce. The variable doing the work was the level of conflict, not the legal status of the marriage.
A 2022 review in Clinical Child and Family Psychology Review (Cao, Fine and Zhou) synthesized the modern literature into what they called the Divorce Process and Child Adaptation Trajectory Typology model. Its core conclusion restates the same point in current language: it is the quality of the predivorce or postdivorce family environment, primarily conflict, that accounts for the variability in children's adjustment, not the legal divorce itself.
Why conflict harms: emotional security theory
The most rigorously tested theoretical framework for explaining how interparental conflict damages children is Emotional Security Theory, developed by Patrick Davies and E. Mark Cummings starting in 1994 and refined steadily since. The core idea is that children monitor their parents' relationship as a primary index of their own felt safety. When that relationship is chronically hostile, children's emotional security is threatened, and that insecurity then drives a range of psychopathology.
The pathway has two steps. First, conflict exposure produces emotional insecurity, which children display through distress, avoidance, behavioral freezing, or attempts to intervene in the conflict. Second, that insecurity, when sustained, produces internalizing disorders (anxiety, depression), externalizing behaviors (aggression, conduct problems), peer problems, and emotion dysregulation. Davies and colleagues have replicated this cascade longitudinally and across multiple samples (Cummings, Davies and colleagues, 2006; Davies et al., 2016, Developmental Psychology).
The 2016 reformulated version of the theory (EST-R) identified four distinct child response profiles, each with its own outcome signature: secure, mobilizing (dramatic distress and appeasement), dominant (hypervigilance and reactive aggression), and demobilizing (freezing, emotional masking, withdrawal). Each predicted unique adjustment outcomes. This is the central scientific point: children process and pay for parental conflict in different ways, but they all pay for it.
A 2008 meta-analysis by Rhoades in Child Development added another layer. Children's responses to interparental conflict show up not just in cognition (self-blame, threat appraisals) and behavior but in their physiology. Conflict registers in their bodies, not just their minds.
The specific tactics that do the worst damage
The literature is precise about which conflict behaviors are most destructive. Four come up repeatedly.
Triangulation
Triangulation is the process by which parents involve a child in their adult conflict. It can take the form of an alliance with one parent against the other, of using the child as a messenger, or of asking the child to take sides. Buehler and Welsh (2009, Journal of Family Psychology), using a four-wave longitudinal sample of 416 families, found that triangulation independently predicted increases in adolescent internalizing problems even after controlling for marital hostility. Bradford and colleagues (2004) replicated the effect across 11 samples from nine countries. A 2024 study in the Journal of Youth and Adolescence(Wang et al.) found that unstable coalitions, stable coalitions, and scapegoating all partially mediated the link between interparental conflict and depressive symptoms in adolescents. Coalitions worked through a loyalty mechanism; scapegoating worked through a guilt mechanism. Different routes to the same kind of harm.
Badmouthing
Rowen and Emery (2018, Family Court Review) published one of the more counterintuitive findings in this area, with the title "Parental Denigration: A Form of Conflict that Typically Backfires." Using multiple informants, the study found that badmouthing the other parent harmed the denigrating parent's own relationship with the child, not just the target parent's. Young adults whose parents had badmouthed the other parent reported lower closeness to the denigrating parent. The harm boomeranged. Rowen extended the finding the following year. The child does not become alienated from the target parent. They lose closeness to the parent doing the damage.
Using children as messengers
Putting children in the position of relaying messages, financial information, or complaints between parents is a form of triangulation that places them in a chronic loyalty bind. Amato and Afifi (2006) found that feeling "caught between" parents predicted lower subjective well-being even in adults at a mean age of 27. The effects of these childhood messenger roles are lasting.
Loyalty binds
Children experience a loyalty bind when they perceive that loving or spending time with one parent is a betrayal of the other. The foundational Buchanan, Maccoby and Dornbusch study (1991) found that about 25 percent of children in high-conflict divorces felt caught in the middle. That feeling predicted anxiety, depressive symptoms, and a reduction in the quality of the child's relationship with both parents.
A note on parental alienation
Any honest treatment of the conflict research has to address parental alienation, because the concept is invoked constantly in family court and because the science around it is sharply contested. The contest is worth understanding.
Richard Gardner coined "Parental Alienation Syndrome" in the early 1980s, based on his own clinical practice. He claimed that 90 percent of children in custody litigation suffered from PAS and that the vast majority of sexual abuse allegations in custody disputes were fabricated by vengeful mothers. The scientific and professional rejection of PAS, as a syndrome, is near-universal. The APA Presidential Task Force on Violence in the Family (1996) wrote that there are no data to support the construct. The DSM-V committee rejected it. The National Council of Juvenile and Family Court Judges has described PAS as a discredited diagnosis. The largest study of child sexual abuse allegations in custody cases (Thoennes and Tjaden, 1990) found such allegations rare and roughly half validated. The claim that mothers fabricate them at scale was not supported by the data.
Distinct from Gardner's syndrome, there is a real behavioral phenomenon that some clinicians describe as alienation: a child's persistent, unreasonable rejection of one parent that is disproportionate to the child's actual experience. Janet Johnston and Matthew Sullivan (2020,Family Court Review) have argued for a differentiated, multi-factor model. Some such cases involve an aligned parent's behavior. Many also involve the rejected parent's own behavior, including abuse or a lack of warmth, and the child's own developmental vulnerabilities. The current scientific consensus is that single-cause "alienation" framings are inadequate, while the underlying behavioral phenomena (denigration, blocking contact) do harm children when they occur.
For parents, the takeaway is simple. Do not badmouth. Do not block contact without a real safety reason. Do not coach a child's statements. Where there are genuine safety concerns, document them clearly and seek appropriate professional and legal support. The fact that the alienation label is sometimes weaponized in court does not change the underlying obligation to keep your own conduct above reproach.
What protects children
The same body of research that documents the harm also documents what buffers against it. Three protective factors come up consistently.
Warm, sensitive parenting
Warm parenting is the single most consistently identified protective factor across this literature. It buffers the emotional insecurity created by conflict, reduces children's self-blame and threat appraisals, and supports the development of emotional regulation. Vélez and colleagues (2011, Child Development) found that authoritative parenting, which combines warmth with structure, directly improved children's active coping with divorce stress over time. Hetherington's Virginia Longitudinal Study identified the quality of the residential parent's parenting as the single strongest predictor of child adjustment after divorce, explaining more variance than income, custody arrangement, or contact frequency with the non-residential parent.
Low-conflict coparenting
Coparenting quality, which is the degree to which parents support each other's parenting and coordinate on the child's needs, is now recognized as distinct from simple conflict measurement and more predictive of outcomes. Bergström and colleagues (2021, BMJ Paediatrics Open), using a sample of 12,845 Swedish three-year-olds, found that once coparenting quality was accounted for, child mental health in different living arrangements after divorce became very similar. Low coparenting quality was the primary driver of elevated problems, more so than the custody arrangement itself.
A related meta-analysis by Krishnakumar and Buehler (2000, Family Relations) confirmed the pathway by which conflict harms children: conflict disrupts parenting itself, increasing hostility and decreasing warmth, and then the degraded parenting harms the child. Conflict reaches the child partly directly and partly through what it does to the parents.
Consistent routines
Selman and Dilworth-Bart's 2024 systematic review in the Journal of Family Theory and Review concluded that family routines (mealtimes, bedtimes, homework schedules) are reliably associated with better behavioral, cognitive, emotional and physical outcomes. They are especially protective during periods of family disruption because they provide predictability and perceived safety, directly countering the unpredictability that conflict creates. Sleep is one specific pathway: Rudd, Holtzworth-Munroe, D'Onofrio and colleagues (2019, Sleep) found that sleep disruption partially mediated the effects of parental separation on child cognitive and behavioral development. Consistent bedtime routines were identified as a protective buffer.
What this asks of parents
The research does not ask parents to like each other. It does not ask them to be friends. It does not even ask them to agree about parenting philosophy. What it asks, in practical terms, is something narrower and much harder.
It asks that the child not see, hear, or feel the conflict. That hostility not happen at the door, on the phone in earshot, or in tense silences the child can sense. It asks that the child not be made into a messenger, a confidant, or a referee. It asks that the other parent not be denigrated in the child's presence, even when the criticism feels accurate, because the cost of that denigration is paid by the child's relationship with both parents, including the one doing the denigrating. It asks that both parents continue to parent: warm, consistent, attentive, present, even on the days the adult situation is unbearable.
This is the hardest work a separated parent has to do. The literature is honest that it is also the most consequential. The variable that the research keeps returning to is, in the end, the one parents have the most control over.
Read next
- What actually makes a co-parenting message court-admissible — Screenshots get challenged. Timestamps get questioned. Here is what family courts actually look for in a digital communication record, and how to make sure the one you bring holds up.
- Children's rights in co-parenting: what they are, and what they are not — Children have a legal and moral right to two functioning parents and a calm environment. Here is what that looks like in practice, and what it does not.
- Parents' rights in family court: what you keep, and what you trade — Filing in family court does not erase your rights as a parent. Here is a plain-language guide to what those rights are, where they bend, and how to protect them.