Thread by @lymanstoneky

I enjoyed helping with this piece for @NCRegister about pets and fertility.

The question here is: do pets replace kids? Or, relatedly, is pet-mania contributing to low fertility?

I argue the answer is mostly no, but occasionally yes. ncregister.com/news/pets-repl…

So to start with, let me note something striking.

Pets are a huge part of many peoples' lives. We spend money on them, we care about them, etc. Human-animal relations writ large are kind of a massive field of human social life and crucial for understanding human society.

And yet, even though shifts related to animal domestication and husbandry are key elements in the rise of settled human life, virtually no social surveys included any questions about animals until very recently. GSS added a pet question in *2018*.

I have been mocked in the past for the fact that in some surveys I have included questions about tattoos and allergies and veganism and all this random stuff, but I genuinely think this stuff matters! Most of our surveys are basically designed to explore economic or class-based

... explanations for social phenomena, or if not then perhaps attitudinal types of explanations.

But it's very rare that surveys explicitly explore what we might call "cultural" differences: what you eat, what holidays you celebrate, how you mark your body, etc.

To the extent culture gets explored, it's almost exclusively through ethnicity, religion, or gender- and family-attitudes of direct interest to demographers. *Occasionally* other topics are surveyed, but broadly speaking surveys about other stuff are very thin on the ground.

So for example, constructing estimates how how many Americans are vegetarian over time is shockingly difficult and you have to use a bunch of single-purpose surveys. Measuring changes in tattoos is practically impossible.

If you want to know how belief in astrology has changed over time, you have a very narrow handful of small public opinion polls until rather recently.

Most survey work just does not explore most of these kinds of questions.

So when a reporter calls you asking, "Are pets replacing kids?" it's often hard to say. You can see two prior attempts I've made at tackling the question here, before I had GSS data on pets: ifstudies.org/blog/fewer-bab… and here with GSS data: ifstudies.org/blog/pet-owner…

And that's why in some recent surveys I've fielded, I have included some questions about pets.

Basically, I ask about pets in two ways:

Most simply I ask "Do you have any pets?" and people can say yes and identify a kind of pet, or say no they have no pets.

The second way I ask about pets is I give respondents 10 "life priorities" and ask them to rank them in order of importance to them. These include things like "good standard of living" or "meaningful career" or "pride in my country" or "time with my family."

And one of those priority options is.... "taking care of pets and animals."

How highly people rate that is a nice measure of *feelings about* animal life. And since I also ask about actual pet ownership, we can explore pet-fertility ties in two different ways.

So, here's how actual number of children ever born varies by women's age and pet ownership.

Mostly, it does not vary much.

On the whole, pet owning women near the end of their reproductive careers actually have somewhat *more* children than other women.

And this points to the first big issue in the pets-family relationship.

Some people might have pets to *substitute* for kids.

But other people get pets to *complement* kids.

Kids like cute animals and ask their parents for puppies!

So when you study pets, "ownership" is really a pretty weird indicator, because *why* people have pets actually varies a lot.

I don't ask people why they have a pet. It's not clear people would be able to give a cogent answer to that question or even what kind of response options to provide.

But my other question about ranking the importance of *care for animals* can clarify this.

Note that pet-owners also have slightly higher *ideal family sizes* as well. So this is not just about having kids; pet owners *want* more kids.

But what if instead of looking at pet-owners, we look at animal-care-prioritizers?

Here, we see that women who rate animal care very highly desire fewer children and intend fewer children.

This negative relationship shows up even if I toss in a bunch of normal controls like age and education and whatnot.

So now we have a paradox.

Pet owners want MORE kids but animal lovers want FEWER kids.

How can we resolve this?

Well, one question is just: do people who value animals more actually have more pets?

In general yes, but note that those numbers are not 100%: there are some people who own pets but don't value animal care highly, and vice versa.

So here it gets a bit complicated, but basically, AMONG PET OWNERS, more animal-priority predicts LOWER desires. Among non-pet-owners, there's no correlation between animal priority and fertility desires.

So, conditional on owning a pet AND on placing a high emphasis on it, that does predict low fertility desires; but note that this is a small group of people. Under 5% of women fall into the "animal-top-priority-and-has-pets" category.

But here's where things get REALLY interesting...

... is that pet priority really what's driving low preferences? Is pet-mania distracting from kids?

Or is it something else?

So remember when I said I survey weird stuff people often make fun of?

Well, I ask people if they agree with the statement, "It would have been better for most people to have never been born."

So for the curious, I do *not* find that this question causes respondent breakoff.

WHICH IS SHOCKING, since asking people if they think human life is basically not worth living is kind of exactly the sort of question you'd expect would tank respondent engagement. Rather, in the ending comments, respondents often flag it as a question they enjoyed answering...

... which just shows you what freaks survey respondents are.

Anyways, do animal-priorities predict this kind of nihilistic view?

Kinda slightly? It's not really very significant, but it seems like maybe high-pet-prioritizers have more negative views of human life?

Again, not highly significant, not least because very few people are at that maximum-priority ranking for pets. But this seems broadly suggestive that maybe pet-priorities are related to generally negative views about human life.

I get similar weakly-positive lines if I model self-expressed worries about population growth, or if I model worries about population growth *exclusive of worries about climate change*.

So, there is a weakly positive link between animal-prioritization and human-pessimism.

I want to emphasize WEAK here. This is not a highly robust relationship.

My view of what is happening here, which I will explore as I get more data with more questions and an expanded sample size, is that for most people, animal-priorities are not predictive of nihilism.

For most people, even high-animal-prioritizers, valuing animal life is non-contradictory with valuing human life; indeed it may be complementary, stemming from a high valuation *of life generally*.

But there exists a *small subpopulation*, whose exact traits I can't quite pin down, for whom intense prioritization of animal life *is* rivalrous with valuation of human life.

This kind of view is quite rare. Under 5% of the population.

And as such getting a significant read on them in surveys requires a big sample size. To find 500 such people to analyze, you'd need to have a sample size of like 10k-50k people. In this sample, I have about 5.6k people. I'll get 3k more in April.

So right now this is all quite tentative and provisional. Pet-priorities do predict lower fertility preferences, pet ownership doesn't, and that's because pet ownership and pet prioritization are not the same thing. Pet prioritization leads to low fertility preferences...

... partly-but-not-entirely because pet prioritization is associated with extremely nihilistic views about the value of human life. However, this group is small and no survey of these issues has yet reached sufficient scale to carefully describe these groups.