What the Research Actually Says, and Why Almost Nobody Tells You
There’s a question that nobody answered me during my divorce.
Why does this hurt more than anyone told me it would?
Not why does it hurt — I expected that. Not whether I caused it — I spent two years on that one, and so does everybody. Not what God thinks — that question was a different conversation, for another blog. The question was simpler and more useful: why this much, for this long, with this shape?
I read the books. The popular ones. The Christian ones. The pop-psych ones. The ones friends handed me with that sad-smile head-tilt people do when they can’t fix it for you. And I noticed something strange. The books were full of prescriptions and full of morals and full of verses. They were not full of research.
So I went looking. I am not a marriage scientist. I am a guy whose nervous system was on fire for two and a half years, and I wanted to know what the published literature actually said — not what the pulpit said, not what the podcast host said, not what the friend-who-survived-it said. What the data said.
Here is what the data said. I wish someone had told me any of this at the time.
What the science actually says
The person whose name you should know is John Gottman. Forty years of laboratory research on couples, the most-cited model of marital dissolution in the field. He found four patterns that predict divorce better than any other combination of variables he tested. He called them the Four Horsemen: contempt, criticism, defensiveness, and stonewalling. Contempt is the worst one. Contempt will end your marriage whether you want it to or not, because contempt tells the other person’s nervous system it is not safe in the room with you. Bodies hear contempt before brains process it.
Gottman also measured heart rate during couple-conflict conversations. When the heart rate crossed roughly 100 beats per minute, conversations stopped being conversations. He called it diffuse physiological arousal. It is what happens when your body decides the person across from you is a threat. You cannot repair a marriage while your cardiovascular system thinks it is in a fight. And nobody taught you — nobody taught any of us — how to notice it in real time.
That is the mild stuff. Here is the heavier finding. Divorce, after it happens, is an HPA-axis event. That means cortisol — the main stress hormone — rises and stays elevated. For months. Sometimes years. Bruce McEwen at Rockefeller coined the term allostatic load to describe the wear and tear of chronic stress on a body. Divorce is allostatic load you can measure in blood draws. When I read this the first time I put the book down and sat on the floor for twenty minutes. Nobody had ever told me that what I was feeling — the half-in-the-world, the can’t-eat, the can’t-remember — was a documented physiological sequence and not a character flaw.
Janice Kiecolt-Glaser at Ohio State studies what marital discord does to the immune system. Slower wound healing. Elevated inflammatory markers. Altered immune response. This is the part where the old script — you’re dramatic, you’re making too much of it, you’re not trusting God enough — runs into the measuring equipment. Your body is not being dramatic. It is running a physiological program that has been observed, replicated, and published.
And then there is grief. We have a whole theology of grief, and we have almost no cultural scaffolding for divorce grief, which is its own species. Margaret Stroebe and Henk Schut at Utrecht proposed a model in 1999 that describes grief as an oscillation — between loss orientation (facing the absence) and restoration orientation (rebuilding daily life). Healthy grief moves between them. The stuck spots are the ones where the oscillation stops.
And then there is Pauline Boss, who gave us the single most useful vocabulary I have ever encountered for this experience: ambiguous loss. Loss without closure. The person is gone but keeps being present — in custody exchanges, in the kids’ stories, in mutual friends’ updates, in photographs you cannot bring yourself to delete. The nervous system needs an endpoint. Divorce refuses to provide one. This is why divorce grief does not finish. It was never supposed to.
The old research on grief taught you to “let go.” The new research — Klass, Silverman, and Nickman’s Continuing Bonds work — overturned that assumption. Survivors do not extinguish attachments. They reorganize them. A twenty-year marriage does not delete. It integrates. You are not broken because you still have feelings about a person you were legally separated from in 2019. You are a mammal with a memory.
What the history actually shows
Here’s the other thing nobody told me. Divorce is not new, and America did not invent it.
Roman law permitted divorce on either party’s initiative from at least the second century BCE. Susan Treggiari’s Roman Marriage walks through the easy-divorce culture of the late Republic. Mosaic and rabbinic law codified a divorce procedure. Indissolubility — the you shall not separate what God has joined reading of Christian marriage — consolidated in the fourth and fifth centuries, largely through Augustine.
The Reformation returned divorce to permission in Protestant lands. Luther treated marriage as a civil matter, not a sacrament. Calvin’s Geneva permitted divorce for adultery and desertion. Henry VIII’s split from Rome was — among other things — about getting a divorce the Pope wouldn’t grant. Christianity has held multiple, contradictory positions on divorce across its history. When a pastor tells you that God hates divorce and the church has always taught this, he is describing a slice of history, not the whole of it. Roger Ross Williams’s documentary work, R.H. Phillips’s Putting Asunder, Stephanie Coontz’s Marriage, a History — all of it is there if you look. Most of us were never told to look.
America had its own arc. Before 1969 you had to prove fault — adultery, cruelty, desertion — and because people who wanted out of bad marriages needed grounds, an entire industry of manufactured evidence grew up. Staged infidelity scenes for photographers. Coached testimony. Migratory divorce — couples traveling to Reno after 1906 because Nevada’s residency rule was six weeks. Glenda Riley documents it; Nelson Manfred Blake wrote a whole book called The Road to Reno. The fault-based era was not a golden age of marital seriousness. It was a perjury machine, and everyone in the legal system knew it.
California passed the Family Law Act of 1969. Ronald Reagan, then governor, signed it — and later said he regretted it. It introduced no-fault divorce to the United States, and within four decades every state had followed. New York was last, in 2010. Whatever you think about no-fault, the argument that America’s divorce rate “exploded because of it” is false. The rate climbed through the 1970s, peaked around 1979-81, and has been declining ever since. Andrew Cherlin at Johns Hopkins has made a career documenting the arc. The Pew Research Center has numbers. You can pull them.
The “50% of marriages end in divorce” line was always a projection, not an observed rate, and has not described any recent cohort. The real picture: the rate is lower than it was forty years ago, especially for college-educated couples, who now marry later, cohabit first, and stay married longer than the 1970s generation.
The one place divorce is rising is midlife. Susan Brown and I-Fen Lin at Bowling Green coined the term gray divorce and tracked its doubling since 1990. That is what S4 Walk-Away Spouses on the primer side is about. If your mental model of divorce is millennials getting divorced, you are reading yesterday’s news. Your parents’ generation is doing most of the current divorcing.
You are inside a long historical arc. You did not cause it. It did not cause you. What this should do — if you let it — is lower the moral stakes of your own divorce enough that your body can do its actual work. Grieve. Reorganize. Reattach. Without carrying a civilizational verdict on top of everything else.
What the impact research actually finds
Now the difficult part. The part where “what about the kids” comes up, and the part where a lot of conservative commentary has been dishonest in one direction and a lot of progressive commentary has been dishonest in the other. Both have cost people who are looking for the truth.
Two studies define the field. Judith Wallerstein followed children of divorce from a clinical sample starting in 1971 and reported — in The Unexpected Legacy of Divorce in 2000 — a sleeper effect: difficulties that didn’t appear in childhood surfaced in young adulthood, especially in the formation of romantic attachments. Her conclusion was that divorce produced lasting, often under-recognized damage.
E. Mavis Hetherington, same era, different approach. Her Virginia Longitudinal Study sampled the community — not a clinical population — and reported in For Better or For Worse (2002) that roughly 75 to 80 percent of children of divorce function in the normal range as adults. The remaining 20 to 25 percent carry elevated risk, but the dominant finding was resilience.
People will tell you these studies contradict each other. They do not. They studied different populations. Wallerstein was looking at families who sought clinical help; Hetherington at families who did not. Both findings are real. Both distributions are real. The truth is not that one was right and the other wrong — it’s that the children of divorce don’t land in one place, and any honest account holds both tails of the distribution.
Paul Amato at Penn State is the empirical middle. His meta-analyses across decades conclude that there are small-to-moderate but statistically reliable effects on children of divorce — academic achievement, psychological well-being, relationship outcomes, self-concept — that persist into adulthood. Smaller than the clinical literature claimed. Larger than the pure-resilience reading implied. Real.
Three findings from the child-outcome literature deserve to be on billboards, because pastors won’t tell you and neither will pop psychology:
High-conflict intact homes produce worse child outcomes than low-conflict divorced homes. Multiple large studies (Amato and Booth 1997; Kelly and Emery 2003). The kids who fare worst are not the kids whose parents divorced. The kids who fare worst are the kids trapped in ongoing parental conflict regardless of the legal status of the marriage. If you were told to stay together “for the kids” inside a house where the conflict never stopped, the research does not support the counsel. The research supports what your body already knew.
Parental conflict after divorce predicts outcomes more than the divorce itself. Constance Ahrons wrote The Good Divorce in 1994 and took criticism for naming the possibility. The research has backed her up. Children do better when their divorced parents achieve functional, low-conflict co-parenting. They do worse when the war continues through them.
Age-at-divorce effects exist but do not determine outcome. Within-cohort variation is wider than between-cohort variation. That means: kids the same age going through the same legal event land in vastly different places, depending on conflict level, co-parenting quality, temperament, support, and a dozen other variables. Your kid is not a statistic. The statistics describe a distribution, not a prophecy.
On the adult side, the Weitzman figure that circulated for decades — women’s household income drops 73 percent post-divorce, men’s rises 42 — was corrected in 1996 by Richard Peterson using Weitzman’s own data. The corrected numbers are roughly a 27 percent decline for custodial women and a 10 percent gain for non-custodial men. Still gendered, still real, still a problem — but nowhere near the original claim. The original numbers are still quoted on the internet. They are wrong. When you see them, you can say so.
What the church got wrong
This section is not theological argument. I am not telling you what God thinks about divorce. That conversation belongs in a different register and I am not running it here. What I am naming is what the research and the history, held honestly, do to the pulpit’s most common scripts. This is clinical and historical observation.
Script one: God hates divorce, and the church has always taught this. The church has not always taught this. The church has held multiple positions across two thousand years and across denominations. Roman Catholic indissolubility is one tradition. Eastern Orthodox economia permitting up to three marriages is another. Reformation-era permission is a third. African Methodist, Pentecostal, and evangelical positions vary widely. When the line is delivered as if it were a unanimous historical position, the line is a pastoral simplification, not a historical fact. You can be faithful to scripture and hold any of several positions. The research showing that some churches have been more punishing than others is consistent and well documented.
Script two: Divorce destroys children. The research does not say this. The research says divorce produces small-to-moderate population-level effects, the distribution contains a resilient majority and a struggling minority, and high parental conflict is a larger driver than the legal event. The destroys framing flattens the distribution into the worst tail, and in some cases it has been used — intentionally or not — to keep people in dangerous marriages. That is not what the data supports.
Script three: Women who leave marriages are walking away from God’s design. This one has the most recent weaponization, particularly through evangelical TikTok. Michele Weiner-Davis coined walk-away wife as a clinical description of a specific pattern: years of chronic unrepaired ruptures, followed by an apparent sudden departure. The pattern is real. Her clinical framing emphasizes the years of unrepaired ruptures part. The weaponized version emphasizes the sudden departure part and skips the decade of emotional groundwork. Weiner-Davis would not recognize the version that’s being preached. I’ve written about this pattern in detail on the primer side — S4 Walk-Away Spouses. Go there for the clinical version.
Script four: If you were a better spouse, the marriage would have survived. Sometimes true. Sometimes false. The research identifies predictors — contempt, DPA, attachment-pairing, chronic unrepaired ruptures — that do not resolve into a single person’s failure. Marriages are systems. Systems can be worked on from one side only up to a point. Past that point, the honest description is not you failed but the system was unworkable and one of you finally named it.
I am not arguing anyone should divorce. I am arguing that the scripts many of us inherited do not reflect what the literature actually says, and that when they are delivered as if they did, they do damage to people who are already carrying a documented physiological load.
What I wish someone had told me
That my body was not being dramatic. That the months of 3 a.m. wakeups, the food that tasted like cardboard, the inability to finish a sentence — those were cortisol. Those were allostatic load. Those were documented.
That ambiguous loss was the word for what I could not close.
That the grief was not supposed to finish. That integration, not deletion, was the shape.
That moving on is a cultural script, not a somatic one.
That the research does not support the you failed God’s design framing my tradition handed me, and that my nervous system was right to distrust the framing even before I had read the research.
That the history of divorce is long, that the church has held multiple positions, and that inhabiting the middle of an arc is not the same as inaugurating a collapse.
That my kids’ outcomes would depend more on conflict level than on whether their parents lived in the same house.
That I could grieve on my body’s timeline instead of the culture’s, and that the body would tell me when the integration had taken hold, without my having to force it.
Where to go next
The primer side of this series has the full clinical version, with citations: S15 Divorce: The Science, the History, the Impact. If you want the body-level detail — what divorce does to the nervous system and why — go to S11 Divorce & the Nervous System. If the midlife walk-away pattern fits your story, go to S4 Walk-Away Spouses.
A short reading list, if you want primary sources:
- Pauline Boss, Ambiguous Loss
- Paul Amato, A Generation at Risk (with Alan Booth)
- Constance Ahrons, The Good Divorce
- Stephanie Coontz, Marriage, a History
- E. Mavis Hetherington & John Kelly, For Better or For Worse
And permission, from one Loopwalker to another:
Grieve on your body’s timeline.
The research is on your side.
Church of NORMAL — Normal Like Peter
“Nothing is lost. Only recompiled.”
