Review: Selfish Reasons to Have More Kids

This is a re­view of Bryan Ca­plan’s book Selfish Rea­sons to Have More Kids. Co-writ­ten with Walid.


Adop­tion stud­ies in­di­cate that differ­ences in par­ent­ing styles have mostly small im­pacts on long term life out­comes of chil­dren, such as hap­piness, in­come, in­tel­li­gence, health, etc.. This means that par­ents can put less effort into par­ent­ing with­out hurt­ing their chil­dren’s fu­tures. If you think kids are neat, then you should con­sider hav­ing more.


Note: We think this is a pretty use­ful book, and it has changed our minds on how many chil­dren we want to have, though nei­ther one us has any chil­dren yet. Also, nei­ther of us are ex­perts on twin or adop­tion stud­ies.

Ca­plan ar­gues that par­ents dras­ti­cally over­es­ti­mate their abil­ity to im­prove the adult lives of their chil­dren. His ar­gu­ment is driven by adop­tion stud­ies, which sug­gest that there is very lit­tle that par­ents can do be­yond tech­niques em­ployed by the av­er­age par­ent that would get them bet­ter re­sults with their chil­dren. Speci­fi­cally, the fol­low­ing ar­eas are iden­ti­fied as ar­eas where differ­ences in par­ent­ing don’t seem to mat­ter:

  • No effect on life ex­pec­tancy, over­all health (as mea­sured by the pres­ence/​ab­sence of par­tic­u­lar health prob­lems and self re­ported health), height, weight or den­tal health.

  • No effect on in­tel­li­gence.

  • No effect on var­i­ous mea­sures of per­son­al­ity: con­scien­tious­ness, agree­able­ness or open­ness (not cer­tain about ex­tro­ver­sion or neu­roti­cism).

  • Lit­tle or no effect on mar­riage, mar­riage satis­fac­tion, di­vorce, or child bear­ing.

But that is not to say that styles out­side of the av­er­age do not mat­ter at all—there are a few ar­eas where par­ent­ing differ­ences do seem to have an effect:

  • A small effect on adult drink­ing, smok­ing and drug prob­lems.

  • A small effect on ed­u­ca­tional at­tain­ment, but no effect on grades in school or on in­come.

  • A large effect on poli­ti­cal and re­li­gious la­bels, such as whether you call your­self demo­crat or re­pub­li­can or Chris­tian or Mus­lim but small effects on ac­tual poli­ti­cal and re­li­gious at­ti­tudes or be­hav­ior.

  • A mod­er­ate effect on when girls start hav­ing sex (but not boys), but no effect on teen preg­nancy or adult sex­ual be­hav­iors.

  • Pos­si­bly a small effect on sex­ual ori­en­ta­tion.

  • A mod­er­ate effect on how chil­dren re­mem­ber and per­ceive their par­ents.

So how do adop­tion stud­ies lead to these con­clu­sions?

Adop­tion stud­ies (If you have a link to a bet­ter overview or dis­cus­sion of adop­tion stud­ies, we’d ap­pre­ci­ate it) help find out the in­fluence of par­ent­ing differ­ences on adult out­comes by com­par­ing adoptees to their adopt­ing fam­ily. If adoptees sys­tem­at­i­cally tend to be more like their adopt­ing fam­ily than like other adoptees along some mea­sure (say re­li­gios­ity or in­come), that im­plies that par­ent­ing differ­ences af­fect that mea­sure.

When an adop­tion study finds that par­ent­ing does not af­fect out­come X, it does not mean that par­ent­ing can­not af­fect it, just that the par­ent­ing styles in the data set did not af­fect it.

The ev­i­dence Ca­plan talks about is pri­mar­ily long run life out­comes. Shorter run life out­comes of­ten do show larger effects from par­ent­ing, but these effects diminish as the time hori­zon in­creases.

If par­ent­ing doesn’t mat­ter, what does?

Ca­plan refer­ences twin stud­ies in show­ing that ge­net­ics have rel­a­tively big effects on all the mea­sures pre­vi­ously men­tioned. This ex­plains why we see strong cor­re­la­tions be­tween par­ents’ traits and chil­dren’s’ traits. He speci­fi­cally uses it to call out at­tributes that we would com­monly as­cribe to par­ent­ing, but may ac­tu­ally have a much larger ge­netic com­po­nent.


Once Ca­plan has ar­gued for the stylized fact that par­ent­ing has only small effects on ma­jor life out­comes, he ex­plores some of its im­pli­ca­tions.

Don’t be a tiger parent

One big im­pli­ca­tion is that you should put less effort into try­ing to make your kids into great adults and more effort into mak­ing your and your kids’ lives more fun right now.

For ex­am­ple, par­ents prob­a­bly spend too much en­ergy con­vinc­ing their chil­dren to eat their veg­eta­bles and learn the pi­ano, given that it won’t af­fect whether they will eat healthy as adults or be more in­tel­li­gent. No one likes fight­ing. If you want your kid to learn the vi­o­lin so they’ll have fun right now, it may very well be worth it, but don’t do it be­cause you think it will in­crease their fu­ture in­come or in­tel­li­gence. If nei­ther you nor your child likes do­ing an ac­tivity, con­sider whether you can stop do­ing it.

Adop­tion stud­ies provide good ev­i­dence that most ac­tivi­ties don’t have a much of a long term effect on your chil­dren, so you need good ev­i­dence to start think­ing that an ac­tivity will be good for your kids fu­ture. The odds are against it.

Have more kids

Fo­cus­ing more on mak­ing your and your chil­dren’s lives more fun means that over­all, hav­ing kids should be more at­trac­tive. If hav­ing an­other kid no longer means fight­ing about finish­ing their broc­coli ev­ery night, maybe it’s not such a bad idea. On the mar­gin, you should con­sider hav­ing more kids. If you were plan­ning to have zero kids, con­sider hav­ing one. If you were plan­ning to have 3 kids, con­sider 4, etc.

Other Topics

In much of the rest of the book Ca­plan gives com­mon sense ad­vice for mak­ing par­ent­ing eas­ier for the par­ents. A cou­ple of these, such as the Fer­ber method for deal­ing with in­fant sleep prob­lems, are em­piri­cally based.

Here are some other top­ics Ca­plan dis­cusses in his book:

  • Hap­piness re­search on par­ent­ing. Ca­plan ar­gues that al­though be­ing a par­ent seems to make peo­ple less happy, the effect is small (Ch 1).

  • Child safety statis­tics. Chil­dren are many times safer than in decades past (Ch 4).

  • Many of the benefits of hav­ing chil­dren come later in life (e.g. hav­ing peo­ple who will come and visit you, etc.), which makes it psy­cholog­i­cally easy to ig­nore these benefits (Ch 5).

  • The ex­ter­nal­ities of chil­dren. He ar­gues that on net, ex­tra peo­ple have large pos­i­tive ex­ter­nal­ities (Ch 6), so you shouldn’t feel guilty for hav­ing more chil­dren.

What parts should I read?

We whole­heart­edly recom­mend read­ing the first 5 chap­ters (121 pages) of Selfish Rea­sons To Have More Kids as these have the most use­ful parts of the book; the rest of the book is less valuable.

Crit­i­cisms of Selfish Rea­son To Have More Kids

There are a num­ber of crit­i­cisms rele­vant to Ca­plan’s arug­ments. For ex­am­ple:

  • Nis­bett claims that hered­ity is much less im­por­tant for IQ than thought (see also coun­ter­claims posted be­low).

  • Will Wilk­in­son claims (one, two) that the cost of par­ent­ing plays a small role in peo­ple’s fam­ily size de­ci­sions, thus it’s not a very strong rea­son to have more kids.

  • Ja­son Col­lins likes the book but would like it to dis­cuss the re­search on non-shared en­vi­ron­ment (i.e. that not ex­plained by ge­netic or par­ent­ing differ­ences, such as peer effects) (link).