Divorce: The Science, the History, the Impact

What the research actually says about the end of a marriage

Chapter S15 · Scenarios · Nervous System Theology · Church of NORMAL

Explainer — Clinical Framing

1. What divorce actually is

Divorce is four things at once, and most cultural conversations about it collapse them together. It is a legal event — a formal filing, a court process, a judgment. It is a relational event — the end of a particular attachment architecture between two people and, often, between multiple extended systems. It is a physiological event — a measurable, long-running disturbance in the bodies of the people involved, including the children. And it is a demographic event — part of a population-level pattern that has a history, a statistical distribution, and a set of predictors.

Keeping these four levels distinct matters. The legal event can end in a single afternoon. The relational and physiological events often take years. The demographic event preceded you and will outlive you. Conflating them produces most of the shame the literature tries to describe honestly and that your body already knows.

2. The science of divorce — predictors and physiology

John Gottman has spent more than four decades building the most-cited predictive model of marital dissolution. His Four Horsemen — contempt, criticism, defensiveness, and stonewalling — emerged from laboratory studies of couples’ interactions, with contempt as the single strongest predictor. Alongside the behavioral markers, Gottman and Robert Levenson documented diffuse physiological arousal (DPA): when heart rate crosses roughly 100 beats per minute during a conflict conversation, information-processing narrows, repair attempts fail, and the conversation stops being a conversation. The marriage is then being run by the nervous system, not the couple.

Marriages also stratify by risk at the attachment level. The combination of an anxious partner pairing with an avoidant partner produces a well-documented pursuit-withdrawal loop (see S2 The Anxious-Avoidant Loop for the scenario-level detail) that, unrepaired over time, is one of the more reliable routes to dissolution. The loop is not a moral problem; it is a regulatory mismatch that the couple cannot solve without help.

Physiologically, separation and divorce are HPA-axis events. Cortisol rises and stays elevated, sometimes for years post-divorce. Bruce McEwen’s work on allostatic load — the cumulative wear of chronic stress — provides the framework: divorce is not one stressor but a cascade (identity reorganization, economic disruption, co-parenting logistics, social-network loss), and the cascade runs a documented physiological course. Janice Kiecolt-Glaser’s immune-function research adds a further layer: marital discord and marital dissolution correlate with slower wound healing, elevated inflammatory markers, and altered immune response.

Debra Umberson (University of Texas) and colleagues, and the broader work of Linda J. Waite and Robert Hughes Jr., document long-tail health effects at the population level — elevated cardiovascular risk, depression, and mortality among divorced adults compared to continuously married peers. The causal picture is complicated (section 7 returns to this), but the observational finding is robust.

3. The science of divorce — grief architecture

Divorce is also a grief event, and the grief literature maps onto it unusually well. Margaret Stroebe and Henk Schut’s Dual Process Model (1999, 2010 updates) describes grief as oscillation between two orientations: a loss orientation (facing the absence) and a restoration orientation (rebuilding daily function). Healthy grief moves between them. Divorce grief often gets stuck in one pole — either chronic loss-immersion or avoidant restoration-only — and the oscillation is what needs restoring, not a timeline.

Pauline Boss’s concept of ambiguous loss is the single most useful vocabulary the divorce reader is usually never given. Ambiguous loss is loss without closure: the person is physically present but psychologically gone, or psychologically present but physically gone. Divorce produces the second kind constantly — an ex-partner who keeps existing, who may still co-parent, who appears at events, whose photographs still exist. The nervous system needs an endpoint; ambiguous loss provides none, and the body stays in open-loop vigilance. S11 explores the somatic consequence; the framework itself is Boss’s.

The Continuing Bonds research (Klass, Silverman, & Nickman, 1996) overturned the older assumption that healthy grief meant “letting go.” What the data actually show is that survivors — of bereavement and, by extension, of divorce — reorganize the bond rather than extinguishing it. A twenty-five-year marriage is not deleted. It is integrated. Cultural scripts that demand “moving on” are demanding something the body does not do.

4. The history of divorce — antecedents

Divorce is not modern, and it is not primarily an American innovation. Roman law permitted divorce on both sides from at least the second century BCE, with no requirement of fault; Susan Treggiari’s Roman Marriage (1991) documents the easy divorce culture of the late Republic and early Empire. Mosaic and rabbinic law (Deut 24) codified a divorce procedure. The early Christian shift toward indissolubility consolidated in the fourth and fifth centuries, with Augustine’s influence central.

The Reformation returned divorce to permission in Protestant territories. Martin Luther treated marriage as a civil matter, not a sacrament, and accepted divorce on limited grounds; John Calvin’s Geneva likewise permitted divorce for adultery and desertion. Henry VIII’s 1530s split from Rome produced the Act of Supremacy largely so the English crown could resolve a marital case outside papal jurisdiction — an episode that illustrates how thoroughly political, legal, and theological the divorce question has always been.

5. The history of divorce — America

The American fault-based era required one party to prove a specific ground — most commonly adultery, cruelty, or desertion. The collateral consequence was widespread manufactured evidence: choreographed adultery scenes, coached testimony, and what historians have called an industry of perjury. Glenda Riley’s Divorce: An American Tradition (1991) and Nelson Manfred Blake’s The Road to Reno (1962) document migratory divorce — couples traveling to Reno (after 1906), then later to Nevada more broadly, where residency requirements were minimal and grounds were liberal.

The California Family Law Act of 1969, signed by Governor Ronald Reagan (a signature he later said he regretted), introduced no-fault divorce to the United States. Within four decades, all fifty states had followed; New York was the last, in 2010. The no-fault revolution did two things at once: it eliminated the perjury industry, and it made divorce practically unilateral, since only one spouse needed to claim “irreconcilable differences.” Stephanie Coontz’s Marriage, a History (2005) argues that no-fault was less a cultural revolution than a legal regularization of what was already happening.

The modern collaborative divorce movement dates from Stu Webb’s 1990 decision in Minnesota to stop litigating divorces and instead work toward negotiated settlement with trained teams. The International Academy of Collaborative Professionals followed; the model now exists in more than twenty countries. Mediation, a parallel track, became widely available in the same era. These pathways have not replaced litigation — most divorces still settle without trial but inside an adversarial framework — but they exist as alternatives for those who can access them.

6. The history of divorce — the arc

The U.S. divorce rate is commonly misdescribed. The “50% of marriages end in divorce” figure was always a projection, not an observed rate, and has not been accurate for any recent cohort. Andrew Cherlin’s The Marriage-Go-Round (2009) and ongoing demographic work (including the Pew Research Center and the National Center for Family and Marriage Research at Bowling Green) document the real arc: the rate climbed from the 1950s through a peak around 1979–1981, and has been declining since — especially for college-educated couples, who are now more likely to marry later, cohabit first, and stay married longer than the 1970s cohort.

The decline is partly real (selection effects — people who choose to marry today are more likely to be those who will stay married) and partly artifactual (rising age at first marriage, declining marriage rates overall, cohabitation absorbing relationships that would once have married and divorced). Susan Brown and I-Fen Lin identified a counter-trend: the gray divorce rate (couples over 50) has roughly doubled since 1990, driven by longer life expectancy, economic independence for women, and midlife re-evaluation. See S4 Walk-Away Spouses / Midlife for the scenario-level detail; here the point is that the “divorce is exploding” narrative is demographically false, while “divorce at midlife is rising” is demographically true.

Internationally, the arc varies widely. Italy legalized divorce in 1970, Spain in 1981, Ireland in 1995 (with second-divorce reforms in 2019); the Philippines as of this writing remains the last country in the world without civil divorce for non-Muslim citizens, though legislation has been under active consideration. Islamic jurisprudence has long recognized multiple forms of divorce (including talaq and khula), with substantial variation across schools and jurisdictions.

7. The impacts of divorce — adults

The adult-outcome literature falls into three buckets: economic, health, and social.

Economic impact is gendered and real. Lenore Weitzman’s The Divorce Revolution (1985) famously reported that women’s household income dropped 73% post-divorce while men’s rose 42% — figures that entered popular culture and political rhetoric. Richard Peterson’s 1996 re-analysis of the same data corrected the numbers to approximately 27% decline for women and 10% increase for men — still a substantial gendered gap, but nowhere near the original claim. The corrected numbers are the defensible cite; the Weitzman figures remain widespread but are statistically wrong. More recent work (Pew, Sayer et al., and Brookings data) broadly confirms the directional finding — custodial parents, overwhelmingly mothers, bear disproportionate economic load — while magnitudes shift with child-support enforcement, labor-force participation, and state policy.

Health impact is observational and contested in its causal direction. The observational finding — divorced adults show elevated rates of depression, cardiovascular disease, and all-cause mortality relative to continuously married peers — is robust across datasets. The causal interpretation is not. The marriage-benefit literature (Waite, Gallagher) argues that marriage itself is protective. The selection critique argues that people who stay married are systematically different — healthier, more economically stable, more temperamentally suited to partnership — and that much of the observed “benefit” is selection rather than causation. The honest stance, per most current reviews, is that both are real: some of the health gap is selection, some is causation, and the ratio varies by outcome and sample. Overclaiming in either direction distorts the research.

Social-network impact is well documented and often under-appreciated by divorcing adults. Rose McDermott, James Fowler, and Nicholas Christakis’s Framingham-based work on divorce contagion (2013) found that divorce spreads through social networks — friends of divorced adults become more likely to divorce themselves — and that divorce produces measurable friendship-network shrinkage on both sides. In-laws are lost. Mutual friends often choose sides or drift. Communities, especially religiously homogeneous ones, sometimes reorganize around one ex-partner. The loneliness is not imagined.

Identity reorganization is the fourth axis. Dan McAdams’s narrative-psychology framework describes the self as an evolving internal story, and divorce typically breaks the story at a structural level — the future you were narrating is no longer the future. McAdams’s work on redemption sequences (reframing loss as precursor to growth) and contamination sequences (reframing an earlier good as ruined by later loss) describes the two shapes narrative reorganization tends to take. Neither is morally superior; both are observable and survivable.

8. The impacts of divorce — children

The child-outcome literature is where honest summary matters most, because bad summary has done real damage in both directions.

Judith Wallerstein’s Children of Divorce project, most visibly in The Unexpected Legacy of Divorce (2000, with Lewis and Blakeslee), followed a clinical sample of children from 1971 forward and reported a sleeper effect: difficulties that did not appear in childhood surfaced in early adulthood, particularly in the formation of romantic attachments. Wallerstein’s conclusion — that divorce produces lasting, often under-recognized damage — is influential, widely cited in conservative commentary, and genuinely grounded in her clinical data.

E. Mavis Hetherington’s Virginia Longitudinal Study, summarized in For Better or For Worse (2002, with John Kelly), followed a non-clinical community sample from 1972 and reported a substantially more optimistic picture: roughly 75–80% of children from divorced families function within normal range in adulthood, though the remaining 20–25% do carry elevated risk of depression, relational difficulty, and academic or occupational underperformance.

These findings are widely described as contradictory. They are not. They are two findings on different populations, using different methodologies — Wallerstein worked with children whose parents sought clinical help around the divorce; Hetherington worked with a community sample. The distribution is real in both studies; the cultural framing around them has been the problem.

Paul R. Amato’s meta-analyses (2000, 2010, and ongoing) are the most defensible empirical synthesis. Amato reports small-to-moderate but statistically reliable effect sizes for children of divorce on measures of academic achievement, psychological well-being, self-concept, and social relations, with the effects persisting into adulthood. The effect sizes are smaller than the clinical literature suggested and larger than the resilience literature implied. Amato’s position, essentially, is that both Wallerstein and Hetherington are pointing at something real; the honest description of the population is a shifted distribution, not a universal wound or a universal resilience.

Several findings are well replicated and worth knowing:

Age-at-divorce matters but not deterministically. Very young children show fewer overt symptoms at the time and more relational effects later; school-age children show the most acute immediate distress; adolescents often show earlier behavioral difficulty but comparable adult outcomes. Within-cohort variation is wider than between-cohort variation.
High-conflict-intact homes produce worse child outcomes than low-conflict-divorced homes. This finding is consistent across multiple large studies (Amato and Booth, 1997; Kelly and Emery, 2003). It is the single most important clinical finding for parents debating whether to stay “for the kids.” The answer depends entirely on conflict level.
Parental conflict post-divorce predicts child outcomes more than the divorce itself. The children who fare worst are those exposed to ongoing high-conflict co-parenting. The children who fare best are those whose parents achieve functional, low-conflict co-parenting — what Constance Ahrons called The Good Divorce (1994).
Parentification is a measurable risk. Children who become emotional infrastructure for a struggling parent — especially an isolated single mother — show elevated rates of later relational difficulty. S11 covers the nervous-system mechanics of parentification in detail.

9. The resilience–damage honest ledger

Most children and most adults come through divorce without lasting psychopathology. A substantial minority do not. The effect sizes are small-to-moderate, not dramatic, and they are population-level statistical findings — they do not predict any given child or any given adult. “My kids are fine” and “my kids were deeply affected” are both possible and both common. The research gives you the shape of the distribution, not the location of your family inside it.

The clinical rule that follows is: take the small-but-real effects seriously without letting them become deterministic prophecy, and take the resilience finding seriously without letting it become minimization.

Normal Like Peter — The NST Section

1. Why the research literature matters for the body

The finding that matters most for readers of this series is the simplest: what you are feeling has been felt by millions of nervous systems, and the shape of that feeling has been documented. Divorce grief is not drama. Post-divorce cortisol elevation is not weakness. The two-year, three-year, five-year arc of reorganization is not “taking too long.” The literature is not a source of judgment; it is a source of permission.

The body is running population-scale physiology inside an individual story. That is what the research is describing.

2. Ambiguous loss and the body’s closure problem

Pauline Boss’s framework is the most useful gift this primer can hand you. The nervous system is a closure-seeking machine — threats want resolution, attachments want presence or clean loss. Divorce provides neither. The ex-partner keeps existing. The co-parent keeps appearing at graduations. The in-laws may still call. The body stays in open-loop vigilance because the loop never closes.

This is why divorce grief does not “finish.” It is not finishable in the shape bereavement grief eventually takes. The work is not to close the loop but to stop demanding the loop close. The grief does not end; it integrates. See H4 Grief & the Nervous System for the broader framework.

3. The sleeper effect and the long tail

Wallerstein’s finding, stripped of its culture-war framing, describes something readers of this series often recognize: old attachment wounds surface later, in new attachments, without conscious cause. A body that organized itself around an absent or conflictual parent in 1983 will run that organization into adult relationships in 2026.

The literature calls this a sleeper effect. The nervous-system reading is that the pattern never slept — it was encoded, it was running the whole time, and adult partnership is the first environment that gave it enough data to surface. See F3 Attachment Theory for the developmental substrate and H5 Inner-Child Debugging for how the pattern sits inside internal-family-systems work.

4. High-conflict versus low-conflict and what your body learned

If you grew up in a high-conflict intact home, your nervous system learned vigilance in a container marketed as safety. If you grew up in a low-conflict divorced home, your nervous system learned two households, two rhythms, and a lower ambient threat load. The research says, clearly and consistently, that the second produces better outcomes than the first — and most people inside both structures were told the opposite.

This is not theology. This is the repeatable empirical finding. Parents who stayed together “for the kids” in a high-conflict frame often produced the very outcomes they were trying to prevent. The honest read is not judgment of those parents; it is permission for the reader to stop carrying shame for a divorce (their own, or their parents’) that the data support as the lower-harm path.

5. Continuing bonds, not moving on

The body does not erase a twenty-five-year attachment. It does not want to. It is not supposed to. Klass, Silverman, and Nickman’s 1996 reframe applies here: healthy post-divorce life is not the absence of the attachment; it is the reorganization of it. The ex-partner does not vanish from the internal system — they take up a different position.

The cultural demand to “move on” is a demand the nervous system cannot meet without pathologizing itself. The research alternative — continuing bonds, integrated, with shifted meaning — is somatically possible. “Moving on” is not a failure you have failed; it is a script your body was right to refuse.

6. The history as permission

If the demographic and historical arc mean anything for the reader, it is this: divorce is not a novel cultural collapse. Rome had it, the Reformation permitted it, the fault-based era produced an industry of perjury, California 1969 was a legal regularization rather than a moral rupture, the 1979 peak has declined since. You are located inside a long arc. You did not cause the arc; the arc did not cause you. What the arc can do is lower the moral stakes of the decision enough that the body can do its actual work — grieve, reorganize, and reattach — without carrying the additional weight of a civilizational verdict.

7. Cross-references

S11 Divorce & the Nervous System — body-level mechanics, stages of divorce grief, parentification, two-household nervous-system load
S4 Walk-Away Spouses / Midlife — the 8-stage deactivation arc and the specific pattern of midlife leaving
S12 Infidelity & the Nervous System — a distinct but adjacent event architecture
F3 Attachment Theory — the developmental substrate that divorce reorganizes
F6 Family Systems — the household as nervous-system environment
F9 CPTSD 101 — how long-arc nervous-system dysregulation compounds
F15 Religious Deconstruction — the frequent co-occurrence and the clinical register in which to hold it
H4 Grief & the Nervous System — the broader grief architecture this primer’s section 2 depends on
H5 Inner-Child Debugging — the parts-work approach to the pre-divorce and post-divorce self
T1 Emotional Regulation — the daily-practice tool set

References & Further Reading

Primary sources — science of divorce

Gottman, J. M., & Silver, N. (1999). The Seven Principles for Making Marriage Work.
Gottman, J. M., & Levenson, R. W. Multiple papers on diffuse physiological arousal in marital conflict (1988–2002).
Johnson, S. M. (2004). The Practice of Emotionally Focused Couple Therapy.
Bowlby, J. (1980). Attachment and Loss, Vol. 3: Loss, Sadness, and Depression.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology.
McEwen, B. S. (1998). Stress, adaptation, and disease: Allostasis and allostatic load. Annals of the New York Academy of Sciences.
Kiecolt-Glaser, J. K., et al. Multiple papers on marital discord, immune function, and wound healing (1990s–2010s).
Hughes, M. E., & Waite, L. J. (2009). Marital biography and health at mid-life. Journal of Health and Social Behavior.
Stroebe, M., & Schut, H. (1999, 2010). The Dual Process Model of Coping with Bereavement.
Boss, P. (1999). Ambiguous Loss: Learning to Live with Unresolved Grief.
Klass, D., Silverman, P. R., & Nickman, S. L. (Eds.). (1996). Continuing Bonds: New Understandings of Grief.
Brown, S. L., & Lin, I.-F. (2012). The gray divorce revolution: Rising divorce among middle-aged and older adults. Journals of Gerontology.
Umberson, D., and colleagues. Multiple papers on marital status and health outcomes.

Primary sources — impact on children and adults

Wallerstein, J. S., Lewis, J. M., & Blakeslee, S. (2000). The Unexpected Legacy of Divorce: A 25-Year Landmark Study.
Hetherington, E. M., & Kelly, J. (2002). For Better or For Worse: Divorce Reconsidered.
Amato, P. R. (2000). The consequences of divorce for adults and children. Journal of Marriage and Family.
Amato, P. R. (2010). Research on divorce: Continuing trends and new developments. Journal of Marriage and Family.
Amato, P. R., & Booth, A. (1997). A Generation at Risk: Growing Up in an Era of Family Upheaval.
Kelly, J. B., & Emery, R. E. (2003). Children’s adjustment following divorce: Risk and resilience perspectives. Family Relations.
Weitzman, L. J. (1985). The Divorce Revolution.
Peterson, R. R. (1996). A re-evaluation of the economic consequences of divorce. American Sociological Review.
Ahrons, C. R. (1994). The Good Divorce.
McDermott, R., Fowler, J. H., & Christakis, N. A. (2013). Breaking up is hard to do, unless everyone else is doing it too: Social network effects on divorce. Social Forces.
McAdams, D. P. (2006). The Redemptive Self: Stories Americans Live By.

Primary sources — history

Phillips, R. (1988). Putting Asunder: A History of Divorce in Western Society.
Riley, G. (1991). Divorce: An American Tradition.
Coontz, S. (2005). Marriage, a History: How Love Conquered Marriage.
Cherlin, A. J. (2009). The Marriage-Go-Round.
Treggiari, S. (1991). Roman Marriage: Iusti Coniuges from the Time of Cicero to the Time of Ulpian.
DiFonzo, J. H. (1997). Beneath the Fault Line: The Popular and Legal Culture of Divorce in Twentieth-Century America.
Blake, N. M. (1962). The Road to Reno: A History of Divorce in the United States.

Clinical educators with accessible public-facing material

John and Julie Gottman — The Gottman Institute (books, workshops, podcasts, clinical training). The most-cited ongoing voice in couple research.
Michele Weiner-Davis — Divorce Busters, The Walk-Away Wife Syndrome. Coined the “walk-away wife” term; her own clinical framing is more careful than the cultural weaponization of the phrase (see S4).
Constance Ahrons — The Good Divorce; low-conflict co-parenting framework.
Esther Perel — Mating in Captivity, The State of Affairs. Her work on infidelity is directly relevant (see S12); her framework on eroticism and long-term partnership shapes preventive reading.
Samantha Rodman Whiten (“Dr. Psych Mom”) — accessible blog-based clinical writing on marriage, divorce, and parenting.
Terrence Real — The New Rules of Marriage; relational-life therapy tradition.

Related primers in this series

Reflection Prompts

Which of the Four Horsemen do I recognize in the marriage I have lived inside — whether in my behavior, my partner’s, or both? What was underneath each pattern?
If I use the vocabulary of ambiguous loss, what in my divorce experience changes shape? What stops being a personal failing and starts being a category the body was always going to struggle with?
Where on the Wallerstein–Hetherington spectrum do I see my own (or my parents’) divorce story? Can I hold both findings as true without collapsing them into a single verdict about myself?
What did the economic impact of the divorce do to my nervous system — separately from what it did to my bank account?
If divorce is a demographic phenomenon as well as a personal one, what does that change about how much of this I carry as personal failure?
What continuing bond remains — with my ex-partner, with the marriage, with the version of myself I was inside it — that I have been told should be gone but is still there?
What part of my divorce story has been shaped by cultural scripts I inherited, and what part is my body’s actual knowing?

Integration Checklist

[ ] I can name at least two predictive findings from the science literature (e.g., contempt, diffuse physiological arousal, attachment pairing) without confusing them with moral judgments about the marriage.
[ ] I understand divorce’s historical arc well enough to locate my story inside a long pattern rather than as an isolated cultural failure.
[ ] I can hold Wallerstein’s and Hetherington’s findings as both true without flattening either.
[ ] I have language for ambiguous loss and can apply it where my grief refuses to close.
[ ] I have distinguished what the literature says about selection versus causation in the marriage-benefit findings.
[ ] I can describe the difference between a child’s experience of a high-conflict intact home and a low-conflict divorced home.
[ ] I know that the economic impact of divorce is real, gendered, and smaller in magnitude than the Weitzman figures that entered popular culture; I cite Peterson’s corrected numbers when the topic arises.
[ ] I have at least one clinical educator from this list whose material I trust for ongoing input.
[ ] I have cross-referenced S11 for the nervous-system side of my own experience.
[ ] I know that moving on is a cultural script, not a somatic or scientific one.

Gentle disclaimer: Normal Like Peter and Church of NORMAL publish trauma-informed educational and creative content. Nothing on this site is medical, mental-health, legal, or crisis advice. If you are in immediate danger or emotional crisis, seek local emergency services. In the U.S., you can call or text 988.

Church of NORMAL — Normal Like Peter “Nothing is lost. Only recompiled.”

Scrupulosity

Healing Cycle