A script for a sixty to ninety second film, written to be read aloud. Every number in it appears in the sections above with its source, so nothing here needs a citation on screen that the page cannot back. The bracketed lines are what appears on screen, not what is said.
[A man's hands. A child's hands.]
A father is not a second adult in the house.
[ON SCREEN: 01 · PSYCHOLOGICAL]
The way he holds a baby at six months predicts that child's language at three. Not how many hours. The range of words he uses.
One father in ten is depressed around the birth, and it peaks between three and six months, when almost nobody is looking.
And fatherhood changes him too. Testosterone falls, most of all in the fathers doing the most care. It is a life stage, not a mood.
[ON SCREEN: 02 · SOCIAL]
Across forty-seven studies built to test cause and not just correlation, losing a father shows up in finishing school, in behaviour, and in mental health that lasts into adulthood.
Across a hundred and nineteen studies of teenagers, weak support from a father predicted delinquency more strongly than weak support from a mother.
These effects are real and they are smaller than the numbers you have seen shared. We use the small ones. They survive.
[ON SCREEN: 03 · RELATIONAL]
The bond is measurable. Rough play, the kind that pushes a child to the edge of what they can handle, is the one parenting behaviour where fathers measure as different.
And when the household ends, distance decides. In tax records for five million children, how far a child ended up living from the parent who left explains twenty-two per cent of divorce's effect on teenage births.
Median distance the year of the divorce: four miles. Ten years later, more than ten.
[ON SCREEN: 04 · SOCIETAL]
Norway reserved four weeks of leave for fathers alone. Uptake went from two per cent to over seventy.
Iceland gave fathers their own month. Divorce fell by up to eight percentage points, and it held for fifteen years.
Maternity leave, world average: twenty-four point seven weeks protected. Paternity: two point two. There is no ILO convention for fathers at all.
And in the two hundred and thirty-four indicators the United Nations uses to measure the world, the word father appears zero times.
[ON SCREEN: it works]
So we already know what works, and we know it is cheap.
Four weeks of leave reserved for fathers alone took uptake from two per cent to over seventy. A month of it in Iceland cut divorce by up to eight percentage points, and the effect was still there fifteen years later.
[ON SCREEN: and then nothing]
Norway ran a full public commission on boys and men. It reported. Its predecessor's flagship proposal drew interest from two municipalities. Its national pupil data register took six years to get a legal basis.
There is still no international convention for fathers. Still no indicator that mentions them. Still no office anywhere whose job they are.
Not because it was tried and it failed. Because after the pilot, nobody was assigned to scale it.
[ON SCREEN: what we will not tell you]
That more hours are proven to fix a childhood. Two countries ran that experiment and the results went opposite ways.
That a father is irreplaceable. One point two million Dutch children say that is too strong.
What we will tell you is this. Where a father exists and wants to be there, how much of his child's life he gets is somebody's decision.
[ON SCREEN: unformentoo.org]
Right now, at the United Nations, that decision belongs to nobody.
Presenting this to a group: the four aspects above are the running order, each section opens with what the evidence can carry and closes with what it cannot, and §06 is written so that a hostile question has already been answered before it is asked. Use the thirty-beat version if you want it on screen instead of read aloud.
The social aspect
Behaviour, school and the world outside the home. This is where the causal designs are strongest and where the honest ceiling matters most: father absence effects survive the rigorous methods, and they are smaller than the numbers that circulate.
This Annual Review synthesized studies using sibling comparisons, individual fixed effects, and other selection-correcting designs. Negative effects of father absence survive these designs but are smaller than in cross-sectional work, and they are concentrated in specific domains: externalizing behaviour, high school graduation, and adult mental health, with much weaker evidence for cognitive ability and adult economic outcomes. This is the honest ceiling for causal claims about father absence.
“strongest and most consistent for outcomes such as high school graduation, children's social-emotional adjustment, and adult mental health”
McLanahan S, Tach L, Schneider D (2013). The Causal Effects of Father Absence. Annual Review of Sociology 39:399-427.FACT 09
Wells and Rankin meta-analyzed 50 studies of broken homes and delinquency (the US DOJ NCJRS index, NCJ 132111, dates the research 1925 to 1985) and concluded that delinquency prevalence is 10 to 15 percent higher in broken homes than in intact homes, with most cross-study variation driven by methodology rather than substance. Association only, no causal design, and the modest size is exactly why the inflated viral statistics fail. The 10-15 percent figure is the paper's own concluding estimate, quoted throughout the scholarly literature (for example Theobald and Farrington 2013); it appears in neither the publisher's truncated public abstract nor the NCJRS abstract, so the verbatim fragment below covers the 50-study design and the methodological finding instead.
“This paper attempts to resolve some of these inconsistencies through an empirical meta-analysis of existing research on broken homes and delinquency. Generally, we conclude that most of the variation in reported effects across the 50 studies is a result of methodological rather than substantive feature”
Wells LE, Rankin JH (1991). Families and Delinquency: A Meta-Analysis of the Impact of Broken Homes. Social Problems 38(1):71-93. DOI 10.2307/800639FACT 10
Hoeve and colleagues meta-analysed 119 samples on parenting and delinquency. Poor paternal support correlated with delinquency at r = -.22 against r = -.14 for poor maternal support, and the difference was concentrated in same-sex parent-child pairs, meaning fathers and sons. These are correlations from mostly cross-sectional studies and the absolute sizes are modest, so the honest reading is that a father's support is not interchangeable with a mother's in this outcome, not that mothers matter less. Publication bias in this literature is documented and the authors discuss it.
Hoeve M, Dubas JS, Eichelsheim VI, van der Laan PH, Smeenk W, Gerris JRM (2009). The relationship between parenting and delinquency: a meta-analysis. Journal of Abnormal Child Psychology 37(6), 749-775FACT 11
Jeynes' 2015 meta-analysis in Urban Education pooled 66 studies on father involvement and academic achievement among urban school children. The overall association was statistically significant and held for both White and minority children. Effect sizes were usually just under 0.2 and are correlational, drawn mostly from cross-sectional and longitudinal survey data, so they show a link rather than proof of cause.
“Paternal involvement, as a whole, yielded effect sizes of usually just under .2 of a standard deviation unit.”
Jeynes WH (2015). A meta-analysis: The relationship between father involvement and student academic achievement. Urban Education 50(4):387-423.FACT 12
Cano, Perales and Baxter used three waves of time-diary data from the Longitudinal Study of Australian Children (n = 3,273 children, ages 4 to 8, Peabody vocabulary outcomes), following the same children across waves, which is stronger than a one-shot correlation. Total father-child time showed at best small associations, and none in the strictest fixed-effects models. Time with fathers in educational activities stayed significant in every model; the authors describe those associations as moderate to large, with 5 extra weekly hours linked to about 9 percent of a standard deviation in vocabulary scores, comparable in their data to the estimated effect of having a university-educated parent. One caution belongs with any time-diary null: diaries usually sample one or two days, which is a noisy measure of a year of parenting, and random measurement error of that kind pushes estimates toward zero, more so under fixed effects, which discard the between-family variance and keep the noisy remainder. A null in this design is consistent both with time not mattering and with time being measured too imprecisely to detect.
“father-child time in educational activities is associated with moderate-to-large improvements”
Cano T, Perales F, Baxter J (2019). A matter of time: Father involvement and child cognitive outcomes. Journal of Marriage and Family 81(1):164-184.FACT 13
Kerr and colleagues followed 206 at-risk American men with annual assessments from ages 12 to 31 (age 26 skipped, 19 waves) and modelled within-person trajectories, so each man is compared with his own earlier path. After the first biological child, self-reported crime showed slope decreases and tobacco and alcohol use showed level decreases; arrests and marijuana use did not change significantly. Drops in crime and alcohol were larger among men who became fathers later, while the decline in tobacco was smaller for later fathers. Single cohort of mostly White, lower-income men from higher-delinquency neighbourhoods: use as an illustration of the settling effect, not as a population estimate. Open full text at PMC3185370.
“Following the birth of a first biological child, men's crime trajectories showed slope decreases, and tobacco and alcohol use trajectories showed level decreases.”
Kerr, D.C.R., Capaldi, D.M., Owen, L.D., Wiesner, M. and Pears, K.C. (2011), Journal of Marriage and Family 73(5): 1101-1116, doi 10.1111/j.1741-3737.2011.00864.xFACT 14