“Starting and not keeping going” is perhaps the most common experience there is. We usually explain it with a single word: willpower. But where does the ability to regulate ourselves come from, and how much does it weigh in life? One of the most cited studies on this question followed the same people for 32 years. It is worth telling properly: for what it says, and for what it does not say.

In short

The Dunedin study (Moffitt and colleagues, PNAS, 2011) followed about 1,000 children from birth to age 32. Among those who had LOW self-control as children, at 32 27 in 100 had more health problems and 43 in 100 a criminal conviction; among those with HIGH self-control, 11 and 13 in 100. Family and IQ explain only part of this difference. But the self-control they measured is not an abstract “willpower”: it is a set of everyday behaviours (impulses, attention, persistence). And the result is an average tendency, not a destiny.

The study

Since 1972 the Dunedin Multidisciplinary Health and Development Study has followed 1,037 people: all the children born in one year in Dunedin, New Zealand. At 32, 96% of them were still taking part, a very rare figure that makes it possible to see the stories of those who had the hardest time too. At 32 they were assessed with medical exams, blood tests, interviews, court records and the views of people who knew them well. In 2011 Terrie Moffitt’s team used these data for a precise question: does self-control in childhood predict how adult life turns out?

How they measured self-control

This is the most interesting point, and the one most often forgotten. Nobody asked the children “how much willpower do you have?”. The researchers combined nine measures collected between the ages of 3 and 11:

  • at 3 and 5, the study’s observers rated the children’s lack of control during testing sessions;
  • from 5 to 11, parents, teachers and the children themselves rated five behaviours:
    • impulsive aggression;
    • hyperactivity;
    • lack of persistence;
    • inattention;
    • impulsivity.

These measures were combined into a single self-control score. In other words, the study does not measure an inner “strength”: it measures how well a child manages impulses, attention and persistence in everyday life, as seen by several people at several moments.

What they found

At 32, those who had less self-control as children were doing worse in almost every area measured: physical health, substance dependence, income, money management, family and the justice system. It did not, however, predict depression.

The numbers, translated

The authors split the children into five groups, from lowest to highest self-control, and compare the two extremes: the fifth with LOW self-control and the fifth with HIGH self-control. Here is what was happening at age 32:

At 32, out of 100 people who as children had… LOW self-control HIGH self-control
…had more health problems 27 11
…were dependent on more substances 10 3
…had a very low income (under 20,000 New Zealand dollars a year) 32 10
…were raising children alone, as single parents* 58 26
…had a criminal conviction 43 13

*Only among those who already had children at 32 (47% of participants).

Put simply: among those with LOW self-control as children, health problems were about 2.5 times as common as among those with HIGH self-control; substance dependence and convictions more than 3 times.

But note: even in the LOW self-control group, most people did not have those problems. 73 in 100 did not have more health problems, 57 in 100 had no convictions. It is a higher risk, not a destiny.

A gradient, not a threshold

There is no line separating “those who have it” from “those who don’t”. All along the scale, a little more self-control went with slightly better outcomes. The link held even when the children with the lowest levels were removed from the analysis, and even when the 61 children diagnosed with ADHD were removed.

How much does the environment matter?

It is the right question, and the study answers it in four ways.

  1. Family matters, but does not explain everything. Children with more self-control came slightly more often from well-off families and had a higher IQ on average. Taking social class and IQ into account, the weight of self-control shrinks but remains. An example: for each “step” down in self-control, the odds of a conviction were 1.8 times higher; taking family and IQ into account, still 1.7 times (technically, an odds ratio of 1.83 and 1.71). For financial problems, self-control weighed more than social class of origin and IQ. Overall, it predicted adult outcomes about as well as social class and IQ.
  2. Between twins raised together. In a second study of 509 pairs of British twins born in 1994–95, same home and same family, the twin with less self-control at age 5 was, at age 12, more likely to smoke, doing worse at school and showing more antisocial behaviour, even at equal IQ. So it is not just “the family you grow up in”.
  3. The “traps” of adolescence. Teenagers with less self-control more often started smoking by 15, left school with no qualifications (more than twice as likely) or became teenage parents without planning to. These traps explain a large part of the link: 32% for health, 42% for crime, 61–63% for single parenthood and substance dependence. But not all of it: even among those who fell into none of the traps, childhood self-control still mattered.
  4. Self-control changes. The link between self-control as a child and as a young adult is only moderate: many people changed position. And those who improved over time had better outcomes at 32. The authors warn, however, that spontaneous change is not the same as change brought about by an intervention.

What the study does not measure is how much of self-control depends on genes and how much on the environment: it only notes that, according to behavioural genetics, both matter.

The limits: what the study does not say

  • Correlation is not destiny. The study is observational, as the authors themselves say: it shows a robust link on average, not that a child who was impulsive is “marked”, nor that increasing self-control automatically produces those results.
  • Self-control does not just mean will. The behaviours measured include hyperactivity and inattention, which also depend on temperament, development and context. Part of what the study calls “self-control” is not a choice.
  • Ratings, not direct measures. The score comes from the judgements of observers, parents, teachers and the children themselves: useful because they differ from one another, but ratings all the same.
  • A specific place and time. The participants were born in one New Zealand city in the early 1970s. The results need confirming in other contexts, and the twin comparison only looks up to age 12.
  • The measures at 3–5 say little on their own. The preschool observations alone did predict adult outcomes, but with modest effects: the full picture comes from all the measures up to age 11 together.

Our conclusion

Self-control matters, and it matters for a long time. But the Dunedin study also tells us what it is made of: impulses, attention, persistence. Not a muscle to strain, but a set of behaviours that also depend on the conditions we live in.

That is why the useful question is not “how do I find more willpower?” but “how do I make what I want to do easier?”: fewer distractions, more concrete plans, people around you. And if you are struggling today, remember two things the study shows with numbers: most people with low self-control as children did not have those problems, and those who improved over time were doing better as adults.

How WeSelf can help

  • Build healthy habits. A goal becomes a week of small actions already in your agenda, with a day and a time: fewer decisions to make when attention is scarce.
  • Find people with your goal. Share a goal with a friend: persistence is easier when someone is expecting you.
  • A community where people support each other. If you skip a day, in the anonymous Community you find people who have been there, without judgement.
  • A specialist, if needed. If impulsivity or inattention really weigh on you, you can ask for a specialist’s contact and you decide what to show them.

An honest note: the study is about psychological mechanisms, not about WeSelf itself. WeSelf does not solve difficult situations: it tries to give a tool to the people living them.

Source

Moffitt T.E., Arseneault L., Belsky D., Dickson N., Hancox R.J., Harrington H., Houts R., Poulton R., Roberts B.W., Ross S., Sears M.R., Thomson W.M., Caspi A. (2011). A gradient of childhood self-control predicts health, wealth, and public safety. Proceedings of the National Academy of Sciences, 108(7), 2693–2698. https://doi.org/10.1073/pnas.1010076108