Bryan Caplan  

The Effect of Children on Happiness: The Latest from the Research Frontier

Repealing the Laws of Supply a... The Financial Regulatory Chess...
Hasty readers of happiness research often conclude that kids are a disaster for happiness.  If you actually look at the size as well as the sign of standard estimates, however, the right conclusion is that kids ever-so-slightly reduce happiness.  I then go on to point out that a great deal of parental unhappiness is unnecessary, so it is reasonable to think that kids combined with scientifically literate parenting would increase happiness.

A recent paper in the Journal of Happiness Research  ("Children and Life Satisfaction" by Luis Angeles) suggests that I'm overly pessimistic.  Contrary to most of the literature, it finds that kids often increase overall life satisfaction.  I was tempted to dismiss this as a typical academic "man bites dog" result, but I was pleasantly surprised when I carefully read the piece. 

The article begins by replicating the standard patterns in a big British data set, the British household panel survey (BHPS): If you run a multiple regression with typical controls, kids slightly reduce happiness.   But then the paper takes advantage of a key feature of the BHPS: It follows the same people over time, so you can re-run the standard equations with fixed effects.  In English, this means that you can see whether individuals' happiness changes when their number of kids changes.

This simple and intuitive adjustment wipes out the standard results.  The data set is so massive (almost 89k observations) that it should be possible to detect even microscopic effects of kids on happiness.  None detected.

To get from this non-result to the "kids increase happiness" punchline, the author makes one last eminently sensible adjustment: It allows the effect of children on happiness to vary by marital status.  This makes a lot of sense - it's a lot easier for a married coupled to raise a child than a single mom.  When you re-run the results this way, you find that as long as they're married, having children makes people happier than they were before.  And contrary to one of my earlier mea culpas, moms gain more than dads.

I have to admit: Based on personal experience, I'm skeptical.  When I run optical fixed effects regressions (="I check whether the people I know seem more or less happy after they have kids"), people almost always seem less happy once they become parents.  But Angeles' paper suggests one way to reconcile research with first-hand experience: Even when kids raise overall life satisfaction, they reduce area-specific satisfaction - particularly for socializing and leisure.  Now consider: Where do I get the chance to observe others' happiness?  In social and leisure situations, naturally!

All things considered, this paper's got a surprisingly plausible story.  I wonder how it will fare against the attacks it is almost sure to provoke.

Comments and Sharing

COMMENTS (9 to date)
arne b writes:

Can something like a genetic predisposition be ruled out as the most important factor in whether children increase ones happiness, i.e. whether it is the initial will to have children that affects how one will be affected by having children?

In other words, if we split the data into four groups
(1) want children, have children
(2) want children, have none
(3) have unwanted children
(4) have none, intentionally
then it is well possible that groups (1) and (4) are both happier than (2) and (3), while the averages of (1)+(3) vs. (2)+(4) are no different.

Sorry if that has been covered before (or in the paper, which I only scanned briefly without finding any clues to it).

Floccina writes:

For me from conception to 4 years old children were negative but since 4 they have been a great delight. Now I wish that I had more.

Robert Speirs writes:

If having children were a truly negative experience, few if any parents would have more than one. That's not the case. Looking at how many parents who've had one child go on to have two, how many with two have three, etc., may give some insight into how parents expect the experience of having children to affect their future happiness.

Taimyoboi writes:

Should a libertarian be an authoritarian when it comes to raising his or her children?

Brian B writes:

Regardless of what studies say, it seems fairly evident to me as a parent and observer of parents that children reduce short-term satisfaction but significantly increase it in the long term. I have little experience with single parents, however, so it is plausible that the stress of single parenthood erases even long-term satisfaction in that case.

Fabio Rojas writes:

I second Floccina's comment. The happiness/children link should vary with age and maybe this is what you are picking with a panel study format. Infants and toddlers are hard people to work with, but you get a good period until the teenage years. I wonder what the BHPS data say about controling for kids age?

Justin Wehr writes:

Some reactions:

Happiness and satisfaction are not the same thing.

Comparing happiness levels before and after kids seems almost completely useless because of the happiness "set point". Consider that people who become paralyzed return to their original happiness level within about a year. (Not to say that having kids is anything like becoming paralyzed, just that major changes in life events have little if any effect on long-term happiness.)

I see happiness as a signal or a by-product of recent events. Studying it as a long-term goal or outcome seems misguided.

eccdogg writes:

I can only speak for myself (a man) and from observation of my wife, but our happiness has gone up even with two young children. A 2.5 year old and a 2 month old (born on the same day as yours Brian).

The toddler is a ton of fun (but also trouble) and there is ablsolutely no substitute for having your child hug you and tell you they love you or having you newborn return a smile.

We really didn't do that much socially anyway and we took your advice and got lots of help.

Steve Roth writes:

Bryan, this is great, and it supports a belief/observation I've often had as the parent of two children:

The second kid makes parenting *much* easier, because only children want your constant and undivided attention. My best piece of evidence is the wonderful line that one mother I knew reported from her child:

"Momma, look at me with your *whole* face!"

Kids are attention sponges/black holes. Two kids provide each other with much of the attention that each craves.

Angeles' study supports this. A jump in life satisfaction (going from negative to positive for the full sample) comes when the second kid arrives. There's another big jump with the third, then satisfaction drops with four or more.

Fixed Effects Full Sample Married Couples
One child -0.027 (0.024) 0.017 (0.030)
Two children 0.009 (0.031) 0.074** (0.036)
Three children 0.059 (0.049) 0.197*** (0.057)
Four or more children -0.007 (0.094) 0.184* (0.105)

(Note that with one exception these married-couple results are the only statistically significant results in Table 4.)

Hypothesis: having two parents provides a sufficient baseline quantity of adult attention (for three or fewer children), at which point the attention provided by one or two siblings goes a long way toward satisfying the the children's (seemingly bottomless) attention needs/wants.

This takes a big load off the parents, increasing their life satisfaction.

With four or more children, even two parents can't provide the baseline level of sufficient adult attention--the other kids' attention doesn't suffice to replace that, and the load lands back on the parents, who can't satisfy it, resulting in reduced life satisfaction.

An old family friend told me and my spouse when we had our first child, "You can't give them too much attention. You can give them the wrong kind of attention, but not too much. Kids act up when they're not getting enough attention."

My experience and observations bear that out in spades.

Another piece of advice based on all this (along with your very wise "get a nanny" suggestion): have lots of play dates.

Comments for this entry have been closed
Return to top