Lyman Stone on Demographic and Marriage Decline

Lyman Stone is a demographer and the director of the Pronatalism Initiative at the Institute for Family Studies. In Lyman’s first appearance on the show, he discusses demographic and marriage decline, the fallacy in the thinking of degrowthers, the benefits of pronatalist policy, and much more. 

Subscribe to David's new Substack: Macroeconomic Policy Nexus

Check out our new AI chatbot: the Macro Musebot!

Read the full episode transcript:

This episode was recorded on February 13th, 2025

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to [email protected].

David Beckworth: Welcome to Macro Musings, where, each week, we pull back the curtain and take a closer look at the most important macroeconomic issues of the past, present, and future. I am your host, David Beckworth, a senior research fellow at the Mercatus Center at George Mason University. I’m glad you decided to join us. 

Our guest today is Lyman Stone. Lyman is a demographer and is the director of the Pronatalism Initiative at the Institute for Family Studies. He is also an affiliated scholar at a number of other institutions, including the American Enterprise Institute. Lyman joins us today to talk about the demographic decline, its economic implications, and prospects for pronatalism policy. Lyman, welcome to the program.

Lyman Stone: It’s really good to be with you.

Beckworth: It’s good to have you on. I have chatted with you for a number of years. In some sense, this is long overdue because this is a topic I do care about. On Macro Musings, we often focus on the business cycle, what the Federal Reserve is doing, what fiscal policy is doing. What’s really important is the long-run trajectory of economic growth.

A key part of that is demographics, population growth, both as an input, but also, what we’ll get to later, as a contributor of ideas and innovation. You got to have people that have those ideas. I’m delighted to have you on, and I know you’ve done a lot of work. It seems like you’re always in the middle of some debate or putting out a new paper. In fact, the paper you put out that prompted this show is a very fascinating one we’ll get to, later, on France. It actually shows that you can make pronatal policies work. The numbers are pretty staggering. I think it was 5 million to 10 million additional people compared to this counterfactual path, had they not done these pronatal policies. 

We’ll come to that later, but let me talk about you for a minute. You are a demographer, you’re in this profession. How does one get started on the path of becoming a demographer and how did you do it?

Lyman Stone’s Career Path

Stone: Well, for me, it was quite accidental. I was originally trained as an economist—that was my undergrad and my master’s—and particularly international trade economics. I was working at a tax policy think tank. I was the new guy. I drew the short straw on topics nobody else wanted to work on. I ended up working on taxes and migration, how taxes influence people’s residential choices.

It ended up just being really interesting to me. When I left that job to work in agricultural economics, I just kept blogging about migration, and then that became not just migration but spatial economics and local area economic change and stuff. There’s also this other part of population change, which is fertility. It’s this other thing that might matter.

Then I started noticing that there were a lot of people interested in just the fertility side. I was interested in it. It seemed like there was big stuff happening there that not a lot of people were paying attention to. This is back in 2016, 2017, 2018. Not a lot of people were really paying attention to this as a major issue. A lot of the research side was not talking to the popular side.

I said, “Okay, I’m just going to start working on this,” and started writing about it more, and then ultimately started working as a market forecaster for companies that make stuff for moms and babies. That’s actually my bread and butter, is I forecast births for companies. Then I also do policy-focused work through the Institute for Family Studies as Pronatalism Initiative.

Beckworth: That reminds me of a story I heard, and I often would share this when I would teach econometrics before I took on the think tank world job. There was a local Target in a neighborhood, and in this neighborhood, there was a father and a teenage daughter. They were at home, and the teenage daughter, in the mail, started getting advertisements for baby products. The father saw this and he was livid. He was mad. He was irate. He ran down to that Target store. He goes, “How dare you question my daughter? She’s not pregnant. She’s not messing around. How dare you?” 

Well, lo and behold, she was pregnant, and Target knew before the father did simply based on buying patterns. They know if a woman would buy certain products a few months prior, they would tend to buy baby products later. Very fascinating story. Very, very illustrative of probably what you have done in your work, those type of patterns.

We want to talk about the broader ones. Before we do that, though, tell me a little bit more about demography as a profession. What do demographers typically do? What are your favorite data sources? What are the hot topics?

Demography as a Discipline

Stone: Demographers, it’s an interesting discipline because you have, I would say, two strands in demography. One is your core classical quantitative demography work where these people are demographers, first and foremost. There’s a set of fundamental relationships of population dynamics that, if you’re in this field, you learn. There are formulas, there are indicators. In the same way that an economist knows, in their sleep, certain accounting identities; a demographer in this tradition knows, in their sleep, certain tempo adjustment patterns for fertility or these different standard tools.

In that sense, it’s a technical discipline. It’s a set of skills and techniques applied to data. It’s not really a set of theories about the world. This central demographic tradition has been associated with stuff like the modernization hypothesis and the demographic transition model. But at its core, it’s essentially a commitment to a set of formulas that you learn; you memorize them, you learn the code for them, and you’re done.

Then there’s a second strand of demography that is going to be things like social demography, economic demography, that sometimes is people who are trained, first and foremost, as demographers, but just as often, is actually somebody trained as an economist or sociologist, or a public health—what do you call somebody who does public health?—a public health scientist, I guess.

It just so happens that, while they’re an economist, they’re interested in labor markets. Family structure influences labor markets. Pretty soon, they’re writing about demography. They present a paper at demographic conferences even though they have no idea what the classical demographic tradition is. They’re like, “We invented this cool new indicator.” Some demographer is like, “Actually, that was invented in 1911. Here’s the formula for it.” Okay.

I had a fun interaction like this with a friend who’s an economist, recently, where he reinvented—it was a bit different than a standard indicator, but very similar. There’s these two different traditions that one is just your classical quantitative demography. The other one is this broader umbrella around it, where, if you’re in one of these traditions, yes, you might learn the central classical demographic tradition, but maybe you also do a whole lot of qualitative interviews to understand people’s ideas about what family means to them. Maybe you are also doing a bunch of models of labor supply elasticity because you’re actually a labor economist.

For me, I would say, in my heart of hearts, I am a classical quantitative demographer. I love my classical proximate determinants of fertility. I think the overall turn in the discipline away from classical quantitative demography and the proximate determinants of democratic demographic outcomes toward this wider penumbra of theories—it’s not that it’s been a mistake. I think that wider range is actually quite important and quite good.

I think we now often have a lot of demographic work being done that is not actually very well informed by relationships that we truly do understand and do know a lot about. Demography is a discipline. We’re one of the oldest social sciences, very clearly. You can go back to John Graunt, in the 17th century, putting together London “Bills of Mortality,” and trying to describe fundamental relationships of life expectancy out of them. This is happening long before you have anything recognizable as economics.

Benjamin Franklin, in the mid-18th century, is already describing the kinds of population dynamics that Malthus is going to describe 30 years later. Franklin does it very clearly in a series of essays about why German immigration is so bad. He might have been wrong on the policy about German immigration, but he did some interesting math. Of course, you have Malthus, who’s like the godfather of economics, all these different things. Demography is, on some level, like the protoscience of social science.

What do demographers care about? Birth, death, and migration. Those are the three. If you’re a demographer, you’re either a baby person, a bodies person, or a moving person. I’m a babies person, although you have to learn some of all of them. I’m a babies person. I specialize in fertility. I like classical demography, but I am really trained. I have an economics background, so I can do the economic stuff. If I’m a coauthor with a bunch of economists, which happens frequently, I can understand what they’re saying. I can keep up, but I allow them to do the actual econometrics. 

In terms of my actual professional expertise, I’m a social demographer, which means I spend a lot of time on survey data. I spend a lot of time thinking about people’s ideas, people’s attitudes, people’s beliefs, and thinking about, “If they have these beliefs and attitudes, how does that walk through to these other parts of the life process that we can measure?”

Beckworth: You’re a bridge between demography and economics. You help—

Stone: In principle, yes.

Beckworth: —remind us, economists, that we’re not the only game in town, which we often think we are.

Stone: I am frequently the demographer, on a list of coauthors, yelling at the economists, “This is not a valid indicator of behavioral parameters.”

Beckworth: Fair enough.

Stone: “We need the actual behavioral parameter here.” 

Beckworth: Economists have to live up to their reputation of being uninformed imperialists, at times. You mentioned Malthus. I read your article we’re going to talk about later. I guess I hadn’t realized that he actually was a key reason for the census being started in the UK, I recall from your paper. He had a big contribution to make to the field of demography himself, not just economics. That was so fascinating to read him and then Benjamin Franklin. Yes, very, very fascinating.

Demographic Decline

Let’s actually talk about that demographic decline. If you go online to Amazon, go to an in-person bookstore, there are a lot of books, some of them very alarmist, talking about a world without people or without babies, or a world without population growth. They really stress, “This is it.” This is quite a big swing from the 1960s when they had the opposite popular message, “Too many people.” We’ll come to this later.

It makes sense that this is in demand now. If you can go to a bookstore and pick up a book on this, then probably there’s an interest and appetite for more. Let’s talk about demographic decline. Maybe you can start us out, Lyman, by talking about it from a global perspective. What is happening in terms of the world’s population growth rate? I understand some countries have absolute declines, not just growth rate declines. Walk us through the big picture here.

Stone: Globally, population growth rates are declining. I don’t recall what the number is right now, but yes, it’s going down. There’s basically only two factors for global population growth—births and deaths—because the number of people who migrate off world is very small, and they tend to return rather quickly.

Beckworth: That we know of.

Stone: That we know of, yes. The alien abductions are pretty demographically significant in Wyoming and Montana, but other than that, yes, we mostly don’t think about it.

Life expectancy increases are continuing apace around the world. People are living longer, being healthier. If that’s the case, if death is coming later and less often, why is population growth declining? The answer is fertility is falling pretty much everywhere. When we talk about changing rate of global population growth, we mostly are talking about falling birth rates.

Birth rates are falling around the world. Because this is happening in many, many places, there is an invidious temptation to look for a single explanation, to say, “This is happening in lots and lots of places, so local explanations can’t be the right answer.” When you actually look at local areas, what you’ll find is the timing, pace, and slope of fertility change really does vary dramatically. There are a lot of countries that have had increases in family size over the modern period.

I want to emphasize I said family size, this is not fertility, because those are actually radically different things. Family size is affected by both births and how many children die because you might just have fewer babies because more of your kids survive, and you were always shooting for the same family size, and you just are shifting the mix of babies that live and babies that die.

There’s, in fact, quite a diffuse trend at the local or national level in this area. There are some big factors like falling child mortality as we’re getting better at keeping kids alive around the world. That is probably the central underlying factor shaping—to the extent there’s a global trend—the extension of modern health innovation, particularly for children, is the central factor suppressing global fertility. Because, when more kids live, people don’t need to have as many live births to hit their desired family size. 

And also, the expected cost of a given conception rises, right? I guess more children can be expected to live. As a child who lives longer is more costly—now they’re also, in the very long run, maybe more beneficial, but people are hyperbolic discounters, and costs are front-loaded, so they’re more costly in a discounted term. People have incentives to restrain fertility as more children live. That’s a central force that you can’t ignore. More kids live, so people have fewer births, but they end up often having more kids, even with fewer births, because of increased survival.

I think that’s important but it’s also boring. It’s actually hugely important for human welfare. It’s great that fewer people are dying. Wonderful, love that. Also, I think, as soon as you say this, people are like, “Oh, yes, okay, I get that.” Child mortality didn’t fall a lot in the US in the last 20 years, but fertility did. When we talk about falling fertility, what people really mean is falling fertility that isn’t explained by this other fundamental, classical demographic dynamic.

Why do we see these fertility declines that are not associated with other fundamental population mechanics? They are the biggest reason. The biggest two reasons are, one, the spread of new ideas, particularly a set of ideas that gets called developmental idealism, and those ideas being linked to smaller desired family sizes. That explains some of the dynamics in middle-income countries, particularly.

Decline of Marriage

Then third, the biggest issue in high-income countries is basically nobody gets married anymore. Nobody’s having sex. Yes, married people have way more sex. Less marriage does mean less sex. It is the case that we’re seeing a massive decline in marriage. In fact, even in a lot of low-income countries, we see this. This is not, not a factor. I just did a piece, a couple days ago, where I showed that half of the fertility decline in India, over the last 30 years, can be explained by declining prevalence of marriage.

Beckworth: Why is that happening? That’s interesting.

Stone: Why is marriage declining?

Beckworth: Let’s start with the United States because I’m sure there’s different answers different places. I have some thoughts in my mind, but I want to hear what the expert has to say. Why is it happening?

Stone: Why is fertility falling in the US over the last 30 years?

Beckworth: Why declining marriage rates as an important part of the story?

Stone: Whenever I say, “Oh, yes, falling fertility is because of declining marriage,” at first, people are like, “Wow, one variable explains so much.” Then they go, “Wait a second, did you just play hide the ball? Why is marriage falling?” You’re like, “Okay, that’s way more complicated.” The first thing to understand is it’s not falling. The immediate thing people immediately jump to is, “Maybe people aren’t getting married as much because they just don’t want kids as much as they used to. Maybe it’s not really that falling marriage is causing falling fertility. Maybe it’s the lower expected future family size jumps backwards through time via an expectational channel and suppresses marriage.”

This is not what’s happening. We actually know this isn’t what’s happening for a variety of reasons. One, we know people’s attitudes toward nonmarital childbearing. Why? Because we’ve been asking about it for decades. We know that people are actually totally fine with it. Second, we know selection into marriage on fertility has declined. How do we know this? Because we know shotgun marriages have declined.

Within marital durations, the period of marital duration where fertility has fallen the most is in the first year of marriage. Right there, that’s telling you massive decline in selectivity of marriage on fertility status. It used to be, you got pregnant, you get married. Now, yes, you don’t need to. You can just have the baby out of wedlock. Interestingly, there’s actually some complicated timing shift dynamics, but I won’t get into that too much.

Furthermore, I won’t belabor the math too much, but there are a variety of mathy ways we can we can explore the selection hypothesis here. My favorite one is just to say, “Look, if we think so, people select into marriage on fertility preferences, let’s just take longitudinal data where we asked people their fertility preferences early on and see if that’s true.” The answer is yes, there is some selection; people who want more kids, get married younger, and spend more years of their life married.

We also see that the gap between people’s stated preferences and their actual outcomes grows the fewer years you spend married. What I mean is, conditional on the number of kids you want to have, you have more kids the more years you spend married, which suggests there’s actually a dose response of marriage within fertility preference status, which is not consistent with a simple selection story. It’s not just selection into marriage based on some expectational channel.

Now, normally, I would skip this whole section of this exploration of this explanation, but this is a podcast for economics nerds, so we have to start with expectational channels here. It’s not that, okay?

Beckworth: Just to be clear, what you said is that, yes, there are some people who get married because they want to have families, but once they get married, they want it even more, that there’s greater expectation, greater desire for more kids once they taste it?

Stone: We can identify people who want kids more or less because we have a lot of surveys where we’ve asked them this, and we have decades of research testing the validity of these surveys, seeing how they predict behavior, seeing how psychometrically credible the measures are. We know they’re quite psychometrically credible, and we know they’re extremely predictive of behavior.

To the extent there are deviations from perfect predictiveness, and of course, there are, we observe that exposure to marriage has a gatekeeping effect on fertility among people of the same fertility preference status. If you say, “I want three kids, and it’s very important to me,” you’re going to have more kids at the end of your life than somebody who says, “I want one kid, and it’s not very important to me.” Those are two very different categories.

If we just look among people who say, “I want three kids, it’s very important to me,” their number of kids varies based on the number of years they spent in a coresidential married union, within group. That kind of dose response—and it is a dose response, that is, the more years of marriage, the more babies—is really not consistent with a simple selection. It’s a large dose response.

Somebody who wanted three kids and it was very important, but never got married, has 0.5 kids or 0.6 kids, okay? Whereas somebody who wants three kids and it was very important, they got married in their early 20s, has three kids. They basically have no gap between their preferences and their outcomes. Marriage has a causal gatekeeping effect on fertility.

The reason is that fertility is a dyadic exercise. It takes two to tango. We don’t need to go over how babies are made. It’s family-friendly podcast, right? It takes two, okay? Why are people not coupling? It’s not because of some downstream decline in desire for children. We have ways to measure desire for children. They haven’t changed very much over time. It’s because of other things.

A big part of it is the declining economic status of young men. On the whole, Americans are getting richer. That’s true. All that decline is like, “Oh, we’re actually getting poorer because middle-class incomes can’t afford washing machines or whatever.” No, Americans are getting richer on average. That’s true. Particularly, American women, and particularly, older Americans.

If you break it out distributionally, you find that the lion’s share of those income gains are from Americans 40 and older, and from women, which I think that’s very easy to understand. If you think returns to human capital are a big part of this story, women do better in school, and the returns from human capital will tend to grow over time, and so you would expect women and older people to do better.

Also, if you think that women’s income was artificially suppressed in the past due to sexism, women now have bigger gains. Then also, obviously, we’ve had a massive expansion in the availability of financial instruments for long-term retirement planning and public subsidy for pensions, right? These are all things that would shift income older, right? This all makes sense.

What that means is that 20-something men actually got a raw deal. The ratio of a 20-something man’s income to a 20-something woman’s has definitely gotten worse. Actually, more importantly is the ratio of a 20-something man’s income to a 40- or 50-something man’s income has fallen a whole lot over the last 30 years. Men in their 20’s have seen zero real economic growth over the last 20 years. Zero, none. Their real incomes have not risen. In fact, they’ve declined slightly.

If you take their incomes relative to older men, they’ve actually declined precipitously. Why does this matter? Basically, when people are coupling, there are certain human universals. Across almost all societies, men like hot women and women like men who have the ability to provide economic insurance. Now, that’s not the only criteria. I’m not saying love or companionship is not a factor, but I’m saying you will not find a society where men are like, “I don’t care about the physical beauty of a mate.”

Whereas you actually do find a lot of societies—and you can see this in revealed preferences, you can see this in dating app data, you can find this all kinds of ways. If you just rate people’s physical attractiveness, it turns out to be highly predictive of a lot of different things. There’s actually a great study that rated a bunch of college students’ physical attractiveness—

Beckworth: I remember this, yes.

Stone: —and followed up with them longitudinally, and found that women rated as more attractive had way more kids, 10 years down the road, but men didn’t, which is relationship success. It turns out how you look is different. I have three kids. I’m not a 10, okay? 

Beckworth: I was thinking you were going to go to the study they did of teacher evaluations at universities, and how they were related to how good they look.

Stone: No. Yes, you can get those too. This isn’t like a gold-digging thing. I actually want to point out, what’s going on here is not the ratio of young women’s incomes to young men’s income. That quantitatively doesn’t matter much. Young women don’t actually care if young men are poor compared to them. Young women care if young men are poor compared to their dads.

I know that that sounds weird. Actually, Greg Clark has a fun paper where he shows that, across hundreds of years of data, women stably match onto men who have the same socioeconomic status as their fathers. There’s zero change—

Beckworth: That’s interesting.

Stone: —in the rate of assortative nature on that, across 300 years of data. The reason is very simple; a young woman’s idea of what a good spouse will provide in their future home is going to be deeply shaped by their observations of what good spouses provided in the generation of their fathers. That might be their own literal father, or it might be the fathers of their peers. Pick your poison on if you think it’s individual socialization or horizontal norms. I don’t care for this.

Sociologists will fight each other over whether it’s family unit socialization or horizontal norms. I don’t care. For the purposes of this debate, it doesn’t matter at the average level for society. The point is that, basically, if young men are really poor compared to old men, young men don’t get married, because they have not yet been able to send the credible signals that they will be able to be what a good spouse is.

A good spouse for men is many things. He might be caring, he might be compassionate, he might be good with the kids, he might do chores at home, or maybe he scalps his enemies and does rain dances, or maybe he has memorized The Analects and honors his father. There’s going to be different cultural norms, but what is cross-culturally the case, is that he provides insurance for his family. That is, women have important subsistence roles. That is, they provide economic goods for the family, whether that is gathering in a hunter-gatherer society, working on the farm in an agricultural society, or in employment in a market-based economy.

Women have always been huge economic providers for their families, but they are providers who have interruptions. Those interruptions occur with children because children create new demands on women’s time and new reductions in their ability to perform work for periods of time. As a result, what women are looking for is men who provide an insurance product. That is, on average, men who are able to insure against the expected volatility of women’s income. Women’s income is, of course, volatile in a way that is correlated with need related to children.

Women experience an income decline at the same time they’ve increased resource needs. Men are an insurance product. They are not providers. This is actually really important to grasp. What women are looking for is not a man to just take care of them, on average. Most women across all societies are actually perfectly capable of taking care of themselves in regular times.

The economic product men provide is insurance. You can also see this in the societies with a lot of violence. Why do you have cultures where women seem to prefer men who have prominent displays of violence? Wouldn’t you think that violent men would be a danger to a woman and her children? The most likely victim for male violence is an intimate partner. Why is it, in a lot of cultures, male violence seems to be mate preferential?

The answer is, if you think of men as insurance products against extreme tail risks in societies where women and their children might be subjected to bridal kidnapping or killing of their children or something like that, men who have effective displays of violence are actually a useful insurance product. I know this sounds very evo-psyche, and I’m not trying to do some crazy evolutionary psychology thing.

At the base level, what I am saying is young men increasingly cannot provide insurance goods. Zoning policies make it impossible for them to buy houses. Houses are the ultimate insurance product, and we know housing access has a massive effect on marriage. Young men don’t have good ways of sending signals that they can provide insurance. Young men are volatile. They look like risky bets.

If you’re a young woman, how do you know that that man is the stability? Sorry, I’m just going on and all. I think men as insurance is actually like a really important heuristic that is not digested enough. This explains falling marriage everywhere. What we’ve seen is skills, bias, technical change has shifted the distribution of income away from young men everywhere.

Beckworth: All right, that’s a key part of the story. You answered the question I was going to ask about housing. Housing seems to be an important problem not just in the US but across many advanced economies. That itself is not the key story. It’s an input to a bigger story about men not able to provide insurance role.

Stone: Housing is one of the mechanisms by which couples insure the family unit. A man’s ability to provide housing is an extremely strong signal of his capacity as an insurer. In principle, really cheap rental housing should be pronatal. In practice, homeownership is quite strongly associated with fertility. If you think of it in budget constraint terms, that doesn’t make sense.

It’s weird that sometimes budgetarily less favorable housing seems to have a stronger fertility association than somewhat more affordable rental housing. It makes sense if you think about this as basically a consumption-smoothing type exercise. Your rent could go up, but a locked-in mortgage, baby, that’s insurance. There’s nothing hotter than good insurance. 

Beckworth: All right. That’s fascinating and definitely an angle I hadn’t thought about. Let’s move on from the causes of this. 

Stone: Sorry, I just have to jump into this for one other weird test of the theory. If this theory is true, then we should expect cohorts that are exposed to more economic volatility to have lower fertility rates. That is, it shouldn’t just be good outcomes, good, bad outcomes, bad. Just bouncy outcomes should be antinatal, okay? It should also be the case that improved financialization, greater extent of financial markets to allow more consumption smoothing should be pronatal.

I will tell you, we actually have good papers with quasi-exogenous shocks to cohort volatility exposures and access to financialization to smooth consumption. Both those families of papers find, yes, greater cohort exposure to volatility leads to lower completed fertility, even if the volatility is upside volatility, and also that greater access to financial tools to smooth consumption leads to higher fertility rates. We actually have tests of the insurance theory of fertility, but it works.

Beckworth: There you go. We need the Federal Reserve to minimize the business cycle altogether. Secondly, we need to continue the growth of the financialization of the US economy, which many criticize, but what you suggest actually would improve prospects of marriage, and therefore, fertility rates.

All right, let’s segue into the consequences of this. I was looking back, when I was preparing for the show, at some of the numbers. In 1800, the fertility rate was about seven per woman. Now it’s under two. It’s a fascinating chart. I’m sure you know this chart well, but if you look at it, it goes down, down, down from the high of seven. It’s getting really low in the 1930s, ’40s. That pops up for the baby boom, which is really fascinating, and then it comes back down and continues on that trajectory.

That baby boom generation, it’s a really fascinating pop. I don’t think we have time to go over the story for it, but I’ll encourage the listeners to check it out. Here we are. We’re on the tail end of that. We’re under replacement, right now, in the US. I believe we’re under 2.1, correct?

Stone: We’re at 1.6 or 1.7.

Degrowthers

Beckworth: Yes, 1.6, 1.7. Many economists are worried. You’re worried. I’m worried. I’ve had people on the show before talk about this. Lyman, there are some people who say, “Rejoice. This is good news.” The degrowthers, they’re very excited about this. Someone we’ve talked about in the show a lot, Paul Ehrlich. He was very famous for his Population Bomb book.

Interestingly, him and others around his time had a bearing on the one-child policy in China. They read that literature. It also had a bearing on our thinking here. Even today, we have these degrowthers that come up. I bring all this up because you had a really fascinating article, a few years ago, titled “The New Malthusians.” Tell us about these new Malthusians. Why are they so eager to keep this birth rate low, and what’s their angle?

Stone: The old Malthusian take was we shouldn’t have babies, because if we have babies, it will have a bad consequence: famine or carbon emissions or depleted zinc deposits, or I don’t know, whatever. The new Malthusian argument is a bit different. We shouldn’t have kids because the future is going to be so bad that those children will be born into an unacceptably miserable state of being. 

There’s a subtle shift here. One of them is a consequentialist argument about the effect of having babies; the other is a consequentialist argument about the quality of life of children born. These are two different sides. The interesting thing is if they’re both correct, then all the people foregoing babies, because of how bad the future will be, would actually make the future better for the people who choose a defection, like the prisoner’s dilemma, right?

Beckworth: Yes.

Stone: All the cooperators who, under the Ehrlich-type old Malthusian argument, who choose not to have kids would actually make the future better for people who choose to defect and to have babies. These are unstable equilibrium-type arguments, but whatever the case, this is the new argument. Ultimately, these arguments are both nonsense. First of all, none of the climate change advisory organizations suggest any population-related interventions, and the reason is we know they don’t work to reduce carbon emissions.

We actually have a great test case where countries suddenly implemented population control policies that were not anticipated and successfully reduced their fertility rates. They did that enough decades ago that we can see what happened to their carbon emissions. The answer is their carbon emissions did not change in trajectory at all. All that happens, if you’re a very poor country, your economic growth rate is almost totally insensitive to your population growth rate.

If you’re Niger or Chad, your population growth rate really doesn’t affect your GDP per capita very much. What affects that is, does a French mining company find gold? At that level, overwhelmingly what we’re talking about is civil wars, diseases, foreign direct investment, portfolio investment, stuff that is just immune to population dynamics. Yes, you can get these really oblique effects, but basically, if you have a civil war, you’re wrecked.

It doesn’t matter if fertility is high or low, if you’re all killing each other. Or if you have natural resources, you get a bonus no matter how many people are in your country. Economic growth is what shapes carbon emissions, it is. Ultimately, what happened is that people got it in their idea that population has a causal effect on emissions, even though it doesn’t, because population has a very complicated causal relationship with economic growth.

If you’re listening, you might catch that it sounds like what I’m saying is population doesn’t affect economic growth at all, and so falling fertility has no economic downsides. Not so fast. Population doesn’t have a lot of effect on economic growth in very low-income societies. But in a society where most of your economic output is basically variations in human capital, well, it turns out that population matters a lot because it alters the stock of human capital.

Among low-income countries, most of it is not huge variations in human capital. As you start to get a more industrialized economy, human capital starts to matter a lot, it does change growth, but it also changes carbon intensity. More human capital-intensive economies also are less carbon-intensive economies. When you’re on the left side of the carbon emissions curve, adding population doesn’t really do anything. In fact, more population probably leads to a less carbon-intensive economy.

Whereas, when you’re on the right side of the carbon-intensity curve, because there’s a classic curve of carbon intensity with respect to GDP per capita, it turns out that adding more population also doesn’t matter that much because you’re getting more and more human capital intense, and so that tends to lead to a lower carbon intensity. In either case, what we actually observe is that population growth rates are shockingly uncorrelated with carbon emissions growth in a standard panel fixed effects–type model, and we can test that with exogenous shocks to fertility policies.

And we can also test this on other things, like deforestation or other types of measures of resource strain, that population just doesn’t really matter that much. What matters is institutional quality, energy mixes, and these other types of things. Degrowthers are just wrong. I don’t know how to convince someone who looks around at the world that, oh, wouldn’t it be better if we all went back in time to when everyone was poorer? No, it wouldn’t. It would be really bad. So many people would be dead. Several of my children would be dead if we didn’t have medical innovations cheaply available that were invented in the last 20 years or 30 years. I don’t want to roll back time, and I think it’s a horrible mistake to do so.

Beckworth: You gave a nice answer. I typically come with this answer, and this would be on the right side of the curve and more of a long-term story. I imagine your answer is probably dealing more with the real-time analysis. To me, people are the solution, not the problem.

Stone: Yes.

Beckworth: Alex Tabarrok has this great saying, “Do you view a person as a stomach or as a brain?” I view a person as a brain, you view them as a brain. For two reasons, I’ll throw out the endogenous growth theory, more people, more ideas, a supply-side story. You’ve got to have so many people before you have an Einstein or Elon Musk or some talented person to help revolutionize.

Stone: Absolutely.

Beckworth: Even on the demand side, the more people, the bigger the market, the more labor specialization. Adam Smith talked about division of labor is limited by the extent of the market. That’s why, right now, we’re dealing with a bit of an anti-immigration push. I’ve heard this from people I know, good friends will say, “If we have fewer immigrants, yes, that’s a negative supply shock, prices go up, but they’re also demanding less, so that’s an offset, right?” And I’m like, “No, because you’ve shrunk the size of the market, you’ve reduced productivity, there’s a real effect.”

Both supply and demand side, people are the solution. This may be long term, may not have a solution overnight, but if you want to solve climate crisis, you’ve got to have smart, innovative ideas.

Stone: I think it’s actually not that long term. First of all, investment decisions are made right now in expectation of the future. If you’re planning an investment that amortizes across 30 years, you care about birth rates right now.

Beckworth: Good point, yes.

Stone: It’s just that simple. Most big bets have fairly long potential payoff times. Secondly, if you look at all these DOGE kids that Elon is hiring into government, you can debate, is this good or bad, but there’s no debating that these young people are productive. Maybe you don’t like how they’re using their productivity, but they are completing lots of tasks very rapidly. I’m not going to say good or bad, it’s not political, but a lot of them are clearly crazy geniuses who could solve really important social problems with the right prompting.

Like this one who solved the burnt scrolls at Pompeii and found a way to read them, people doing this are often like 18 or 19 or 20. Fertility rates started really falling in the US about 16, 17, 18 years ago. It’s amazing the number of people who I can remember them saying, “Fertility is a long-run problem. It’s not urgent, it’s not pressing.”

Now I see them freaking out in articles in Chronicle of Higher Education being like, “We’re only a few years out from a cliff in enrollment at universities, and no one knows why.” I’m like, “If the human lifespan is about 80 years, a problem 20 years down the road is not far away. I will experience that problem for half of my life.

Beckworth: Yes, good point.

Stone: It blows my mind when people are like, “It won’t matter for 30 years,” and I’m like, “That’s really soon.” Do you realize how soon 30 years is? That’s so close. Investment decisions right now depend on 30-year forecasts. 

100% agree with you. People are brains, and yes, scale generates you more Einsteins, and ideas scale super efficiently. Anton Howes has an awesome piece out yesterday about industrialization and spread of ideas on how to use coal from Germany to England, and it’s very clear that population scale is part of that story. On the whole, I just totally agree.

I think we should really push back on the idea that these are long-term future problems. Falling future growth is a problem in this exact moment. It is destroying investments today. I want to tell you, I know this because I have special insider knowledge and that special insider knowledge is I work with pharmaceutical companies, food companies, consumer product companies, who make decisions about, should we build a new factory or should we close a factory? Should we invest in a special type of baby formula that’s only going to be used by people with a rare medical condition, and it’s going to be $350 million of research, or should we not invest in it?

Should we develop a vaccine for a rare virus that causes childhood blindness or should we not? That calculation depends on how many babies are born. My industry is uniquely exposed, I will grant that, but it really is the case that investment decisions are made right now, affecting many hundreds of thousands or millions of American jobs that are about birth rates right now.

Possible Policy Solutions

Beckworth: That is a nice segue into the next part of our discussion, and that is, what can we do about this? If this is a problem now, what do we do about it? Can government policy do something about it? Lyman, most of the conversations I’ve had, a lot of what I’ve read will say, yes, this is a problem, but pronatal policies are expensive and the return on your investments, very low, small bang for your buck.

I was scrolling through X and I found some tweets that you had, and you have this article titled, “Does Pronatal Policy Work? It Did in France.” And, man, you had some pretty striking claims in there. Tell us about those claims, and then how could we apply that to the United States?

Stone: The basic story is this. Starting in the 1920s after World War I—actually, I got to go back farther. France had one of the first fertility transitions in the Western world. Their fertility started falling in the mid to early 1700s. They had super low fertility compared to the rest of Europe, all through the 1800s and up through World War I. This caused critical manpower shortages during World War I against Germany, which had had a later fertility transition and all these extra soldiers.

After World War I, French strategists realized this was a big problem and they decided to go pronatal. They started rolling out policies in the 1920s, 1930s, but they really got generous after World War II, because after World War II, they were like, okay, we really—never again. They rolled out these really, really generous policies, and they kept rolling out more and more extragenerous policies for the next 60, 70 years. Those policies have been studied in at least four different economic publications, I know of, studying six different policy interventions.

They vary a bit in their judgments of effectiveness, but on average, they imply—again, they’re not all directly measuring total fertility rates because you know how quasi-experimental studies are. Sometimes you can’t directly measure—you’re measuring one parity progression or something. If you walk them out to what they imply about fertility elasticities, they suggest that French pronatal policies boosted fertility by about 0.1 or 0.2 children per woman on average.

I published a paper last year where I showed new evidence, one on border discontinuities over time. I showed that over the course of time when France was rolling out these policies, their fertility rates shifted from about 0.1 kids below their neighbors along border regions, to about 0.1 above their neighbors, or about 0.2 above their neighbors, which is consistent with these things. Also, I was able to show that this is not because of immigration; that native-born French women have about 0.2 children more on average than other similar women around Europe.

That’s true even if we just limit to French women who have four French grandparents. I also showed that it’s not because French women want more kids. The gap between desired and actual fertility is about 0.2 kids smaller for French women than women in other countries. This all says France raised fertility by about 0.1 or 0.2 kids per woman. That sounds small. That’s the type of effect that we summarize when we’re writing about this. When I’m writing about this for all policymakers, I say, there are modest effects of pronatal policies.

That’s always what I’ve said and that’s true, because that’s not a huge change. If you’re a country like South Korea and your fertility rate is 0.6 or 0.7 children per woman, and you’re like, this fertility will increase me to 0.8, that’s still very far below two. People are like, it won’t work, it won’t get us back to replacement. True. What if instead of saying how far are we from replacement, we just asked how many extra people did this policy add over 80 years, because it’s been 80 years since these policies were implemented? That’s about three generations.

The answer, in France’s case, is about 5 to 10 million extra French people exist today because of pronatal policies. Those French people, they are French people who have four French grandparents, overwhelmingly. Back-of-the-envelope math from relative fertility rates suggest something like 80% of those extra people have four French grandparents and the remaining 20% have at least one grandparent who is in France, have at least one French grandparent, that is, they’re not fully immigrants. This is overwhelmingly a shift in the native stock for France.

For comparison, the Paris metro area has, I think, 12 million people, and the Paris municipality has 2 million. We’re basically saying without pronatal policy, France wouldn’t have Paris.

Beckworth: That’s amazing. It’s large.

Stone: That’s massive.

Beckworth: Yes.

Stone: When I first did this calculation, I was like, am I nuts? Am I massively overestimating the elasticity of population with respect to fertility? The answer is no. A quick and dirty way to test this is to go look at the UN’s population forecasts, and they produce ranges. They say population will be this, this is our forecast, but here’s what it’d be if fertility was 0.5 kids higher, here’s what it’d be if kids were 0.5 lower. They basically produce the same estimated elasticity of future population with respect to fertility rates.

Depending on the country and your assumptions about mortality growth, migration, and stuff like that, basically 0.1 higher fertility rates increases population 80 years later by between 2% and 20%. Again, your opinions about how migration and mortality respond are going to shape your view of where you fall in the 2% to 20% range. Usually, the range is going to be in like 5% to 10% per 0.1 kids added. That’s actually a huge effect when you think about it in terms of an 80-year time span, which an 80-year time span is basically one person’s lifespan.

Across one human lifespan, pronatal policies that are modestly effective actually compound; it’s one human lifespan, but it compounds across two or three generations and it gives you millions and millions of extra people. In the US, we would be talking about tens of millions of extra people. We’d be talking about adding California or something like that—or maybe not California, adding Illinois or New York.

I think that this should really calibrate the way we think about effectiveness. If instead of saying, oh, but we’re still so far below replacement, we said, how many million people are we adding, counterfactually? It’s a very different kind of calculation.

Beckworth: Yes. The key to this story then is to make a little change, but keep at it for several decades, that’s the key, and that’s the power of compounding.

Stone: Policy durability is everything.

Beckworth: It’s just a small change in the growth rate, just a teeny bit.

Stone: This is what I find is, economists often look at this and they’re like, but that effect is so small. It’s like they turn off the exponentiation function in their brain when they’re thinking about fertility. And I’m like, don’t you understand? It’s compound. Twenty, 25, 30 years down the road, you get more. And then they’re like, it’s only a tempo effect. It pulls births earlier. If you shorten the compounding period, if the bank charges you the same interest rate, but they charge it every six months instead of every year, that’s a radically different long-term interest rate.

It doesn’t matter if it’s just a tempo effect on some level. It’s the same thing. If you just change the compounding time, it’s as good as changing the interest rate, which is also like, it’s the hack that if you can convince your bank to let you pay your mortgage every two weeks, instead of every month, you save actually a ton of money over the lifespan of your mortgage because of the compounding cycle and et cetera.

Beckworth: Yes.

Stone: Once you realize that TFRs are compound indicators, you suddenly realize, oh, if you can maintain pronatal policy, it’s actually gets cheaper the longer it lasts, so to speak. That does mean the political calculus has to be about maintaining it and that is problematic, because if you think of like a hyperpartisan government coming in and being like, we’re going to do pronatalism. You’re all going to have babies now, maybe the next party will repeal it. If you do pronatalism primarily by executive order, that is easy to repeal. What you need is constitutional amendment pronatalism.

Beckworth: Good luck with that.

Stone: This is what I’m saying is like, it does complicate the policy calculus because you want to do pronatalism, but not in a way that makes half the country decide to go and get sterilized in response. I really hope that the Trump administration does stuff and actually, they should do stuff by executive order. This Department of Transportation order they did recently, I’m 100% in favor of it because I think it’s better to do something rather than nothing.

If you were to ask me, do I think the implied elasticity of these policies is going to be on the higher or lower end of the published literature, I would guess a little more toward the lower end, not because the administration is doing something stupid or bad, they’re not, but because the polarized political environment we have may make it challenging to do anything because some people will have adverse responses. I think that’s also why we should be thinking about nonstate-led pronatalism.

That is, are there things that just individuals can do to alter the birth rates of other people? I have a paper that’s going to be coming out in a few weeks in the Journal of Population Economics where we show that, yes, nonstate actors or parastate actors, in our case, can massively shape fertility behaviors. We look at a religious intervention in the country of Georgia using three different quasi-experimental methods. We use difference and difference synthetic control and interrupted time series on three different datasets identifying this intervention. We show that, yes, a nonstate actor boosted fertility in Georgia by about 0.5 kids, and that sustained and it’s been going for 15 years, and Georgia is still way above its pre-intervention trajectory. 

This, to us, suggests that cultural engagement is really important, too. It’s not just about having policies; it’s about those of us who think that babies are great and people should have them talking really forthrightly and publicly about how great babies are, and maybe offering to babysit our neighbor’s kids or offering to babysit our child’s kids, our grandchildren. I always like to say that if you have grandchildren and you’re not babysitting them, you really can’t say you think low fertility is a problem.

Beckworth: Boom, okay. We are near the end of our program, and I will note here that Lyman, you actually have an article with some suggestions for what we could do in the US that would outlast one president. We’ll provide a link to that article, but I’ll just briefly list the four things you mentioned: Fix the EITC’s anti-marriage structure, improve child tax credit structure, stop discriminating against stay-at-home parents, and fix perverse incentives in retirement programs. We’ll provide a link to that in the transcript. With that, our time is up. Our guest today has been Lyman Stone. Lyman, thank you so much for coming on the program.

Stone: My pleasure.

Beckworth: Macro Musings is produced by the Mercatus Center at George Mason University. Dive deeper into our research at mercatus.org/monetary policy. You can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. If you like this podcast, please consider giving us a rating and leaving a review. This helps other thoughtful people like you find the show. Find me on Twitter @DavidBeckworth and follow the show @Macro_Musings.