Field of Science

Why do atheists have fewer kids?

Here's something interesting from the papers last week. First we've got the philosopher Julian Baggini, an atheist, arguing in The Guardian for the virtues of a childless life. Then, in response, Ed West writes in The Telegraph arguing that atheism is facing a kind of demographic implosion, as the religious inexorably overwhelm them in in the fertility arms race.

Now, this isn't a new argument, but there's precious little research into it. About the only sociologist brave enough to attempt a quantitative prediction is Eric Kaufmann - I blogged about his latest analysis back in June.

But it does give me a handy hook to talk about a chapter in the recent book The Biological Evolution of Religious Mind and Behaviour by Michael Blume - who's done a number of studies into the function of religion from an evolutionary perspective.

First off, some basic stats to give you a feel for what we're talking about here. These are averages across all nations in the World Values Survey, showing the tight light between fertility and religious service attendance.

So is this simply because religions are associated with traditional values? Or maybe that the religious are lower socio-economic status.

It doesn't seem to be so. Blume zooms in to Switzerland, and the data from the census in 2002. Those Jewish and Christian sects that have a higher proportion of the wealthy and educated are actually more fertile than the others.

What's more, traditional sects seem to have lower fertility than the new ones, like Jehova's Witnesses and the New Apostolic Church.

It seems, then, that there is a direct effect of religion on fertility. The question is why that might be so.

From an evolutionary standpoint, it's clear that any trait that increases reproductive success will become more common in the gene pool. Assuming that the demographics we see in the modern translate into the modern world, those genes that favour religion would be more successful.

It's not at all clear to me that we can extrapolate back like that. After all, there's a lot more to reproductive success than churning out children. And modern people have retained a capacity for atheism, which suggests some competing reproductive benefit.

Still, it's worth considering why religion is linked to higher fertility. And it's here that Blume's arguments get really interesting. He suggests that a key factor is honest signalling.

This is the idea that people pay a visible price to get membership of a group, in order to prove that they are committed to the group. The classic example is initiation rites in gang membership.

How does that apply to religion? Well, religions impose a number of obligations on their member - service attendance, food and dress codes, for example. The idea is that these obligations deter those who are not 'true' believers.

There are a number of issues with honest signalling theory as it applies to religion - the jury is still out in the matter (personally, I'm sceptical). But Blume does provide one tantalising piece of evidence.

And that's gender ratios in religious membership. Specifically, the heavy preponderance of women in religious groups, followed next by married men.

The idea is that women have a lot to lose by hooking up with an unfaithful guy. But a male who has made a commitment to the group is sending a signal that he values the group ideals sufficiently to invest the time and effort in going to Church (or whatever). With a bit of luck, that means he's not going to run off with the next available female that crosses his path.

If you want to read that essay yourself, you can download it from Blume's webpage here. It's worth it for the fascinating anecdotes about Darwin and the splendid Faust reference!

Creative Commons LicenseThis work by Tom Rees is licensed under a Creative Commons Attribution-Share Alike 2.0 UK: England & Wales License.


  1. And modern people have retained a capacity for atheism, which suggests some competing reproductive benefit.

    I would be very skeptical of the idea that a 'capacity for atheism' must have a reproductive benefit. You basically saying that just because there are atheists around, this capacity must be selected for. That does not need to be so. Besides, if the numbers are really true, and assuming religious belief is heritable, then it is hard to argue that a capacity for atheism increases fitness.

    However, I would also doubt that religiosity is heritable enough that it can overcome 'cultural evolution,' meaning that even though religious people have more children, atheism might still grow when the children are influenced by factors that cause them to lose their faith (education, in the broadest sense).

  2. Like the previous commenter, I suspect this has very little to do with genetic selection and much more to do with "memetic" selection.

    If you have one group of people that believes that contraception and family planning is wrong, and another that does not believe this, which do you think might have more offspring, all else equal?

  3. It doesn't surprise me that rational people are also responsible when it comes to over-population...

  4. @Andi
    Why do you think all religious people are anti contraception and anti family planning.It varies for different religions and different sects within a religion.
    Have you not heard of Atheist fertility expert who wanted to keep multiple copies of his evolved genes in the pool? He impregnated many women with his sperms till he was caught.That much is enough to speak about responsibility!

    General comments:

    Higher fertility could be due to not resorting to abortion in case of failure of contraception among the religious groups.

    Higher fertility could be evidence of more stable marital relationships among the religious.

    The idea of honest signaling theory makes heuristic sense but has exceptions in cases of religious minorities.For example if I a Hindu who becomes a Christian in India,I lose opportunities for arranged marriage at home.I lose opportunities to marry ones from Christian background despite my faith because of my family background.My conversion surely signals honesty.Why would I become a Christian when there is possibility of persecution?It must be because of genuine faith.But this reduces chances of reproductive success.Isn't it?

    I hope atheists don't join religious groups for reproductive success.They know that the girl would not go away with the next guy who comes along.They are welcome for other genuine reasons :-)

  5. Toim,

    I would be very interested to know what you think of the book as a whole. The thing that strikes me is that I did not recognise any one of the authors as a cog sci of religion person even though they have a lot to say on the topics in the book. I do know that Steadman & Palmer's as well as Sosis' work comes from a somewhat different direction.

  6. Bjorn, arguably, religion results from a combination of cognitive biases to do with agency detection etc, coupled with social group bonding psychology. Now, it's easy enough to imagine a world in which these are so heavily selected for that any notion of atheism becomes impossible.

    In other words, even starting from the assumption that there are no gods in reality, there's no a priori reason to suppose that we should have the mental capacity to discover that.

    Of course, we do have that mental capacity. Which suggests that that mental capacity - whatever it is - provides some reproductive advantage.

    Of course, it may not. It may just be a byproduct :)

    But you could imagine that too much religion is positively harmful. So you would end up with a mean (in absence of cultural effects, a predisposition to be somewhat religious) and a distribution about that mean.

  7. Konrad, I haven't read the book - Michael sent me his chapter. At £63, I'm not likely to either! I haven't come across many of the authors before, although I have seen some stuff by Rossano (and Steadman + Palmer & Sosis). They're sociologists - but McNamara is a psychologist. The book does look interesting. I'll have to see if I can persuade my local library to get hold of a copy.

  8. I was just wondering why the title of the post asks "Why atheists have fewer children?" and then discusses why religious people have more children?

  9. Dear Tom,

    thanks for the fair and insightful blogpost! After we have had ("tested") related studies and discussions in Germany for some years, I was eager to know if there would be people out there interested in an evolutionary understanding of religiosity and religions. Thus, I began to do some translations and original works in English, planning to publish some more in the future.

    As to some opinions formulated: Personally, I would agree that religiosity started as a by-product of "normal" cognitive processes, quickly evolving since then. Although we have some hints linking religiosity of hunters and gatherers or the religious competition i.e. of the late Roman Empire to fertility, I would assume that the effects are today stronger than ever - we are living amidst an observable process.

    After having made very good experiences in the German blogosphere, I would love to share thoughts, ideas and studies with you and other interested people abroad. I just twittered this post (as @Evoreligiosity) and I am looking forward to reviewing your fascinating study in my new English Scilog "Biology of Religion" and hope we'll have more chances to focus on our shared interest. Personally, I believe that the Internet is offering new and distinct possibilities for interdisciplinary and international sciences.

    Best wishes to you and all the contributors to your fascinating blog!


  10. From the perspective of Reproduction and Fertility in medicine, it is highly implausible that people of religious faith are scientifically more fertile than atheists. Surely that claim cannot be made or substantiated without extensive medical research, which to my knowledge have not yet been carried out.

    It cannot be denied that many religions frown upon the use of contraception and birth control and sometimes even encourage large families. In some underdeveloped nations, birth control is rarely available and often unreliable due to a general lack of education. There are also mixed messages on its use due to religious intervention, regardless of intention.

    Furthermore, the term 'religious' is so vague in this arguement that it would be impossible to make such a brash assumption without breaking the subjects up into a greater number of more accurate categories. The parameters are presently too wide.

  11. Dear @jonesy,

    thanks for your inquiries. Well, we tested religiosity from different perspectives, especially practices (as frequency of prayer or worship attendance) and religious affiliation. We used studies and data from diverse censusses. And of course, as you may have read, we checked for other factors as i.e. education or income. If you are interested, you can find the Web Resources on Religion and Reproduction on my homepage, listing ranges of respective studies from different scientists. And the list is growing, with conclusive results.

    What's more, we explored case studies as the Amish, Hutterites, Mormons or orthodox Jews, whose high birth rates cannot be explained without their religious affiliation! (The same is true i.e. of the Shakers, who endorsed childlessness and thus quickly exited (bio-)cultural evolution. You cannot explain their behavior without their religious beliefs, either.) The potential of religions to influence behavior and outcomes in the field of reproduction can no longer be denied seriously.

    With many colleagues in the evolutionary studies of religion, I define religiosity as "behavior towards supernatural agents" (that is: agents whose existence cannot be verified scientifically as i.e. God, gods, angels, space aliens, demons, spirits, ancestors). It is oberservable and explorable in the framework of evolutionary sciences as any other behavior. That's what we do.

    Best wishes!

  12. Thank you for your rapid response, Michael. It seems more apparent having read your reply that the issue I have rests of a possible difference in the definition of the term 'fertility'. Reproduction and fertility are related fields, but cannot be classified under the same specialised heading.

  13. @jonesy

    I clearly agree with you on that one! Differences in understandings and definitions are very frequent in this interdisciplinary field! In evolutionary studies, "fertility rates" are classical terms comparing the numbers of surviving offspring from different populations of phenotypes. But with regard to humans, we are normally used to understand the same term in more medical contexts: I.e. as the "ability" of people to have children they may want to have ("to be fertile"). Therefore, I try to concentrate on the terms of "reproduction", "reproductive benefits / advantages" etc., although they might sound rather mechanical. But it will take time to develop shared understandings of words. You might just imagine the fray of sociologists, biologists, historians, demographers and psychologists cooperating and debating, using terms as "religion", "success", "God", "gods", "nature" and "culture" etc., often having very different meanings in mind! Thereby, all progress in this field is relying on questioning each other with tolerance and patience! :-)

    Therefore, thanks for your insightful inquiries!

  14. Tom, you say "From an evolutionary standpoint, it's clear that any trait that increases reproductive success will become more common in the gene pool." And that is clear, over generations.

    But the loss of faith, at least in Europe, and the growth in new religions movements seem both to be much faster than evolution could explain. So the numbers of fast-breeding believers probably depend more on cultural factors than on genetics.

    That said, it may not apply within communities that are isolated from the mainstream. Within UK Jewry the orthodox are outbreeding the liberals I believe.

  15. David, atheism is clearly cultural/learned. Like a scientific understanding of aerodynamics, it's not selected for genetically (at least, that can't explain the rise in atheism in recent decades).

    But to understand aerodynamics requires certain fundamental intellectual capabilities - mathematics etc. What I'm saying is that it's not axiomatic that we should have evolved this capacity. The only reason we can learn aerodynamics is because a capacity for abstract reasoning and logic was useful in our evolutionary past.

    One reason that people can adopt atheism is that we can overcome our tendency to 'overdetect' agency. This ability was evolved because it brought reproductive advantages.

    In other words, our inbuilt capacity for atheism exists because it increases reproductive fitness.

    Having said that, the balance between religion and atheism is determined culturally, at least in the timescales we're talking about. The question is, how will the balance between conversion and reproduction pan out in future generations?

  16. Tom, I fully agree with your answer to David. Although we Homo sapiens may suppress distinct aspects of our religiosity as i.e. overdetection of agency, we can never get fully rid of it, as shown by a range of psychological experiments on adults and children. For example, even (and especially) ardent secularists repeatedly started to venerate deceased main figures in mausoleums and shrines, simulating their omnipresence by pictures and statues, citing their scriptures, memorating them on special days etc.

    Nevertheless, a range of findings (i.e. from international surveys and twin studies) indicate that there's a broad individual variation in religiosity, as it is in musicality, intelligence, language use and any other biocultural traits. We'll find sceptically inclined and deeply religious among all populations, with the overall percentages of believers fluctuating due to historical, political and socioeconomical circumstances (if better off, the needs and benefits of religious affiliation tend to sink, whereas the opportunity costs tend to raise -> more seculars).

    Nevertheless, there has never been any modern population without percentages clinging to religious behavior - not even China or North Korea could stamp it out. And if given the liberty, the devout members of older and new religions tend to reproduce on higher margins (or to exit the process). Therefore, I can't see how cultural evolution could "beat" biology in the field of religion(s) - as i.e. in the late Roman empire, religious minorities will tend to regroup and regrow again and again. (Biocultural) Evolution is a process, and we're a part of it.

  17. Tom,
    Well I have to agree that "our inbuilt capacity for atheism exists because it increases reproductive fitness" but I wouldn't put it that way.

    I think we have at least two relevant abilities and one attitude which, taken together, give us the capacity for atheism:
    1. The ability to reason abstractly.
    2. The ability to reflect on our own thoughts and feelings.
    3. The sense that we should sometimes prefer the results of abstract reasoning to our more spontaneous thoughts and feelings.

    Each of these is the result of evolution but whether each separately or this particular combination provided increased reproductive fitness seems far less clear. One or more might be an accidental byproduct or might increase fitness only in combination with some other ability.

  18. just an idea to talk about.
    does having fewer children result in having smarter children? if so, by having fewer children does it lead to having atheist children due to them being smarter? is that a possiblity?
    just bringing up discussion.

  19. That's certainly one explanation put forward for the Flynn effect - and there is some evidence that smaller family = more intelligent kids (e.g. here

  20. Dear Anonymous, dear Tom,

    although I would agree that smaller families might invest more in each of the offspring, there are some exceptions in the field of religion(s). Think of those Jewish communities combining high numbers of children with higher education and IQ-levels. The same is true for some smaller Christian denominations adapted to high-educational-environments (i.e. Australian Lutherans etc.). These communities usually combine binding religious life (and thus: bigger families) with strong commitment to education and preference of intelligent mates.

    Best wishes!

  21. Hi Tom,

    at the Explaining-Religion-conference at Bristol University, the topic of the reproductive advantage has been presented and debated.

    Susan Blackmore wrote an impressive piece about that, and concerning our good discussions, I would love to read some of your thoughts about that:

    Best wishes!


  22. Micahel, yes I saw that and thought, "sigh..." I mean, higher fertility in the modern environment doesn't mean that religion has a fertility advantage under other conditions, still less an evolutionary advantage. Non-religion has grown, showing that ideas can spread. We know ideas can spread. Religious priming has all sorts of complicated effects. Cultural evolution is well understood, and religion isn't a genetic product. In short, although 'memes' were always a theoretically dodgy construct, it seems a peculiarly naive view that she's suddenly espousing. If she got that from going to the 'explaining religion' conference then either the conference wasn't up to much or she wasn't paying attention!

  23. She also seems to finish by saying that 'memes' can only be harmful. Which seems a bizarre idea. If she genuinely thinks that, that religion then no wonder she's a bit confused. If she previously thought that religion was all harm and no good, then it's a good thing she's been disabused. As we all know, it's a whole heap more complex than that!

  24. Those are all very good questions, Tom, and I've asked most of them to Michael. Most definitely his evidence for traditional societies was much weaker than for modern ones. I think most people, including Sue, appreciated that the demographic story is just one element of the whole picture but Michael's figures were pretty astounding.


  25. Atheists that I know, are very concerned with overpopulation and consumption. These concerns lead most of them to chose not to reproduce.

  26. Having lived in Afghanistan for almost five years, I can assure you that unless you're willing to produce at least seven to ten children and survive with almost nothing, you'll eventually be outbred or genetically assimilated. Countries that allow immigration, encourage birth control, and discourage parenthood gradually become diluted with foreign DNA. I'm an atheist and I intend to have at least five children one day, but I've already lost a generation because I'm 36 years old.

    Enjoy your first world life before it's diluted and destroyed. But if you truly want your genetic material to survive, I recommend moving to Africa. You can't have it both ways.

  27. Obviously Blume has never heard of Islam which, in most countries, is almost exclusively a male religion. This seems to work strongly against Blume's assumption of "the heavy preponderance of women in religious groups".

    1. Wes James its not Blume who hasn't heard of things but you and your Christian club. Your sentence is meaningless, what does that even mean "exclusively a male religion"?

      How could Muslims reproduce if they are exclusively a male religion?

      Most white converts to Islam are actually women not men, read up.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS