The relationship: how big the change really is, and where the 67% came from
Two meta-analyses, thirteen years apart, agree. The average decline in the first year is small, and it is the same size for mothers and for fathers.
Something does change, and it has been measured carefully enough that you can be told how much.
The larger analysis pooled 49 studies, 106 separate samples and 145,139 participants. Effects are reported as a standardised size, written g. By the usual convention, 0.2 counts as a small effect, 0.5 as medium and 0.8 as large.
| Change in relationship satisfaction | Mothers | Fathers |
|---|---|---|
| Pregnancy to 12 months | g = -0.31 (95% CI -0.38 to -0.25) | g = -0.29 (-0.36 to -0.22) |
| 12 months to 24 months | g = -0.16 (-0.24 to -0.12) | g = -0.14 (-0.27 to -0.02) |
A second, older meta-analysis of 37 studies following couples across a first birth, with nearly 6,000 participants, found g = -0.27 for women and -0.23 for men. Two independent teams, thirteen years apart, dozens of different studies, and essentially the same answer. That is about as solid as this field gets.
These are small effects. That sentence needs a note, because the 2022 paper prints the convention itself, in its own words: We used Cohen's (1988) recommendation regarding the magnitude of effect sizes, namely 0.2 for a small effect, 0.5 for a medium effect and 0.8 for a large effect
. Two sentences later the same paper describes its own values of -0.31, -0.29 and -0.28 as All the effect sizes were medium.
By the rule they had just set out, they are not. This guide gives you the numbers and calls them small, and tells you that the authors labelled them otherwise, so you can judge for yourself.
It is not mostly her experience. Whether the person was the mother or the father did not change the size of the decline in either year. Nor did age. Nor did how long the couple had been together. If one of you is bracing for this to happen to her, the evidence says it happens to both of you, by the same amount.
An average hides a very wide spread. In the 2009 analysis, the individual study results ran from -0.99 to +0.27 for men and from -0.93 to +0.17 for women. Some studies found relationship satisfaction going up. Its authors concluded that becoming a parent may not be a universal stressor, or 'crisis,' for couples, as early theories suggested
.
Some of the decline was going to happen anyway. Both analyses compared parents with couples who did not have a baby over the same period. The 2022 analysis found non-parents declining too, but by less: g = -0.12 for women and -0.13 for men. The 2009 analysis found parents in its comparison studies declining by g = -0.22, almost identical to the -0.23 it found for parents overall. The two do not agree about how much of the drop belongs to the baby, and it is more honest to say so than to pick one.
We discovered that 67% of couples experienced a precipitous decline in relationship satisfaction in the first 3 years of the baby's life. The same authors' 2005 paper puts it as a range, 40% to 67%, and credits their own year-2000 study.
That study followed 130 newlywed couples for six years. 43 of them became parents. Its published abstract reports which things predicted a stable relationship and which predicted a declining one. It does not report that proportion at all. So a figure quoted everywhere as though it described the population traces back to 43 couples in one American city, in a paper that does not appear to state it. This guide quotes it above only to show where it came from, and does not repeat it as a finding.
Two other things you may be offered, which do not exist. No study found for this guide reports what share of first-time-parent couples got worse, stayed the same, or improved. Meta-analyses pool averages; they do not count couples. And no study reports how much more new parents argue. Figures like "couples argue eight times more after a baby" have no traceable source.
"What would be the sign, for you, that we need help rather than more time?"