Are Smartphones Causing the Global Fertility Plunge?

I was lying in bed watching stand-up comedy when I came across a new comedian from this year, Duoduo, talking about how much he loves playing on his phone. The bit was genuinely funny. He said, “I’m going to put playing on my phone on my résumé. What’s wrong with playing on my phone?”

That reminded me of an issue that has received concentrated attention from mainstream Western media over the past two months: To what extent might the spread of smartphones have driven a new round of falling fertility rates worldwide?

In May 2026, the Financial Times published a lengthy data report titled “Why Are Fertility Rates Falling Everywhere at the Same Time?” On June 8, The New York Times published “Why Is Fertility Falling? Two New Studies Point to Phones,” introducing two recent studies that examined data from the United States and from around the world, respectively. The reports appeared less than a month apart, and both pointed to a relationship between smartphones and low fertility worldwide.

What had remained a controversial hypothesis in academia has now entered mainstream public discussion in the West. The Financial Times and The New York Times both have considerable influence in international news, economic reporting, and public-policy debate. The author of the Financial Times report, John Burn-Murdoch, is the paper’s chief data reporter. He has long conducted cross-national data analysis on population, public health, and social trends, and is known for statistical visualization and comparative research. His article brought together population records, time-use surveys, housing data, internet-search trends, and the approximate dates when fertility rates began to turn in different countries.

The New York Times report centered on two new economics studies that examined whether identifiable structural changes appeared in US and global fertility rates after smartphones entered everyday life.

One point needs to be clear at the outset: These reports did not portray smartphones as the sole cause of low fertility worldwide. Housing prices, income, employment stability, childcare costs, women’s educational attainment, the institution of marriage, gender relations, and people’s confidence in the future all remain important factors in understanding changes in fertility.

The new research addresses a more specific question: Why, beginning around 2007, did fertility rates enter a highly synchronized, accelerating decline in countries with enormous differences in their levels of economic development, welfare systems, religious traditions, and family cultures?

I want to explore four questions through these two reports. How does the global decline in fertility since 2007 differ from the long-term decline of the past two hundred years? What methods do the two latest studies use to identify evidence of smartphones’ influence, and how strong a conclusion can that evidence support? How does the decline in face-to-face socializing ultimately affect the number of births through dating, sexual relationships, and partner formation? Finally, if part of the root of contemporary low fertility lies in a decline in social interaction, why do government subsidies, tax reductions, and expanded childcare often have only limited effects?

Two centuries of fertility decline

We need to distinguish between two historical processes. The first is the long-term fertility transition experienced by modern societies over the past two hundred years. The second is the new round of decline that began around 2007.

If we trace the US total fertility-rate curve all the way back to 1800, we can see a very clear long-term trend. At the beginning of the nineteenth century, the average American woman had approximately seven children over her lifetime. After that, the number fell by another step every thirty or forty years, from seven to six, then to five and four. By around 1900, the US total fertility rate had already fallen to between three and four. A baby boom occurred after the end of World War II, but the long-term direction did not change. Today, the US total fertility rate is already below two.

The gap between seven and two looks astonishing, but an important qualification is necessary: Infant and child mortality in the nineteenth century was far higher than it is today. A woman having seven children in the past did not mean that all seven would survive to adulthood. If we calculate the number of adult offspring each woman ultimately had, the figure around 1800 may have been only about three.

Part of the fertility decline during the early modern era came from the reduction in child mortality. In the past, families had to have many children to ensure that some survived. As public health, nutrition, vaccines, and medical care improved, parents could be more confident that their children would grow up, so families no longer needed to have large numbers of children to maintain their number of descendants.

The economic structure also changed fundamentally. In agricultural societies, children could begin working at a very young age. They could look after livestock, help with farmwork, and do household chores. Children were both family members who needed support and a source of labor and old-age security.

Industrialization and urbanization completely changed that calculation. People moved into cities, living spaces became smaller, child labor was restricted, and the amount of time children spent in education kept increasing. Parents had to invest more time and money in their children before they reached adulthood. Children gradually changed from participants in the family’s productive activities into a long-term child-rearing responsibility requiring sustained investment.

Changes in women’s status were also extremely important. Women spent more years in education, entered the labor market in greater numbers, and gained more economic opportunities and social choices. As a result, the time required for childbearing and childcare carried a higher opportunity cost. Developments in contraception also allowed women and families to control the timing and number of births more effectively.

Other factors included urbanization, secularization, individualism, changes in the institution of marriage, and a whole series of additional causes. These factors emerged at different times in different countries. The fertility transition generally occurred first in high-income countries such as those in Western Europe and North America, then spread to middle-income countries, and then to lower-income countries. It was a relatively slow, phased process of modernization.

We therefore cannot attribute the entire decline in fertility over the past two hundred years to phones simply because the iPhone debuted in 2007. The smartphone theory is trying to explain one recent phase within a long-term trend.

By the end of the twentieth century and the beginning of the twenty-first, fertility declines had gradually begun to slow in many developed countries. US fertility was relatively stable in the 1990s and early 2000s. The formerly very steep declines in some middle-income countries also began to flatten, and a few European countries even experienced modest rebounds.

These changes once led demographers to reach a relatively optimistic conclusion. After fertility had fallen dramatically during the initial stages of modernization, they wondered whether it might stabilize near the population-replacement level. Wealth, education, and women’s employment had indeed reduced fertility in the initial stage. But demographers proposed that, once societies grew richer and their welfare systems, childcare services, and family policies gradually improved, people might be able to rebalance work and family, allowing fertility rates to recover to some degree.

John Burn-Murdoch proposed a broader explanation. He said the decline in fertility in Western societies became more gradual during the period from the fall of the Berlin Wall, German reunification, and the collapse of the Soviet Union through to the 2008 financial crisis because the West experienced a relatively stable era. The risk of international war seemed to be declining, economies continued to grow, and globalization kept expanding. In the West, this period was often described as the era of the “end of history.”

The atmosphere of an era is difficult to prove with a simple statistical model, but it may still affect people’s confidence in the future. Deciding to have a child means linking the next ten or even several decades of your life to another life that is highly dependent on you. People’s assessments of social stability, economic prospects, and the future order all enter into that decision. When a society broadly believes that tomorrow will probably be better than today, people are more willing to take on long-term responsibilities.

After 2007, this relative stability was shattered in the West. The financial crisis, long-term economic stagnation, unaffordable housing, political radicalization, wars, the pandemic, and climate anxiety appeared one after another. In surveys, young people have repeatedly said that they do not know what kind of world they would be bringing a child into.

Yet these explanations cannot fully account for the characteristics of the latest downturn. This new decline displays an unprecedented degree of global synchronization. Beginning at somewhat different points between roughly 2007 and 2015, fertility rates in different countries began to fall at an accelerating pace once again.

There is no need to be overly precise about the exact year because smartphones and social media did not suddenly enter everyone’s life on one particular day. A new generation of the iPhone debuted in 2007, but it took several years for smartphones to go from expensive consumer products owned by a minority to routine tools of everyday life among young people. Countries also differed in their mobile-internet infrastructure and phone prices.

What is truly noteworthy is the international scope of the change. Societies including the United States, Britain, Australia, South Korea, Mexico, Indonesia, Senegal, Egypt, and Iran differ enormously in their political systems, religious traditions, economic conditions, family structures, and welfare policies. Yet they developed fertility-rate curves with similar shapes.

Of the world’s 195 countries, more than two-thirds now have total fertility rates below the 2.1 needed to keep their populations stable over the long term. In 66 countries, average fertility is now closer to one than to two. In some countries, the most common number of children a woman has over her lifetime has fallen to zero.

Low fertility used to be concentrated mainly in wealthy countries such as Japan, South Korea, Italy, and Germany. It has now spread rapidly to middle-income countries. In 2023, Mexico’s fertility rate fell below that of the United States for the first time. Brazil, Tunisia, Iran, and Sri Lanka subsequently entered a similar range. Many countries may have to begin confronting population aging, a shrinking labor force, and rising pension burdens before they have become wealthy.

The speed of the decline has also repeatedly exceeded the forecasts of population models. Only five years before 2023, the United Nations predicted that approximately 350,000 babies would be born in South Korea in 2023. The actual number was only 230,000, meaning the forecast was about half again as high as the reality. Demographic change is generally regarded as slow, stable, and easy to predict. But realities in recent years have shown that previously constructed models may have underestimated the speed at which reproductive behavior can change.

Why the financial crisis is not enough

The year 2007 was immediately followed by the 2008 global financial crisis, so an economic downturn may seem like the most direct explanation. During periods of economic difficulty, unemployment rises, household incomes decline, and young people postpone marriage and childbearing. This is very common historically.

The financial crisis may indeed have triggered the initial declines in the United States and some European countries, but the economic explanation faces three difficulties.

The first concerns timing. The US economy later returned to growth, the labor market gradually improved, and stock prices and other asset prices rose sharply, yet fertility did not rebound with the economic recovery. Research in the Nordic countries has also found that the financial crisis may explain short-term changes in 2007 and 2008, but has difficulty explaining why fertility continued to fall in 2018, 2022, and 2026.

The second difficulty comes from cross-national comparisons. Australia is relatively similar to the United States and Britain in culture, language, and income level, but it was affected far less severely by the global financial crisis. Nevertheless, Australia’s fertility rate underwent a similar turning point. Some countries in the Middle East and Southeast Asia maintained fairly robust economic growth during this period, yet their fertility rates also fell at an accelerating pace.

The third difficulty concerns policies and institutions. The Nordic countries have more comprehensive social-security systems, parental leave, childcare services, and gender-equality policies. A relatively high share of young people there also live independently. Yet fertility in these countries has continued to fall in recent years.

This does not mean that economic conditions are unrelated to fertility. Housing stability is especially important. The Financial Times analysis concluded that, in some countries, as much as half of the fertility decline since the 1990s may be associated with lower homeownership rates among young people and an increasing share living with their parents.

A lack of stable, long-term housing obstructs other long-term commitments. If someone cannot say where they will be living next year, it is difficult for them to plan for marriage and children. However, changes in housing and income are generally relatively gradual. They cannot fully explain the sudden acceleration of the past decade or more, or the common shift spanning the entire world.

Demographers needed to find a new factor that had simultaneously emerged in the United States, Europe, the Middle East, Latin America, Southeast Asia, and parts of Africa. This brought the development of digital technology into view.

One important argument put forward by the Financial Times is that fertility declines in the past and the present are occurring at different stages of the family-formation process. In the past, fertility fell mainly because married couples had fewer children. A family that might once have had four children later had only two. Families were still formed and marriage was still widespread; the change was in the number of children born after marriage.

Over the past decade or more, an increasingly important shift has been the growing number of people who do not form partnerships or families. Data from the United States and most high-income countries show that the average number of children born to women who become mothers has generally remained stable, and in some places has even increased slightly. The most obvious increase has been in the proportion of women who never have any children, as well as the proportion of people who never enter marriages or stable cohabiting relationships.

If marriage and cohabitation rates in the United States had remained at their levels from a decade ago, today’s total fertility rate might be higher than it was ten years ago. In other words, the critical point in the current US fertility decline is gradually shifting away from how many children each couple has and toward how many people manage to become couples in the first place.

This changes the starting point for policy analysis. Childcare subsidies can reduce the cost to an existing family of raising a second child, but they cannot directly help a person who has been single for years and rarely participates in in-person social activities to find a partner. Maternity leave can reduce the burden on working mothers, but it cannot solve the problem of young people never entering intimate relationships at all. Tax reductions can affect couples who have already begun planning for children, but they are of limited help to people who have not yet formed a partnership.

Low fertility is therefore increasingly coming to resemble a crisis of family formation. The precondition for a family-formation crisis is a decline in socializing, dating, and intimate relationships.

The iPhone as a quasi-natural experiment

The two studies covered by The New York Times approach the issue on two levels: one within the United States and the other through global comparison.

The first study was conducted by Middlebury College economists Caitlin Myers and Ezekiel Hooper. Its title was somewhat provocative, roughly meaning, “Is the iPhone a Contraceptive?” The study took advantage of the iPhone’s unusual early sales model in the United States.

From June 2007 to February 2011, Apple had an exclusive partnership with AT&T. During that period, American consumers who wanted to use an iPhone generally had to connect to AT&T’s telecommunications network. But AT&T’s mobile-broadband coverage was not the same everywhere in the United States. Some counties had almost comprehensive coverage, while others had very little. Yes, that’s right: The United States really does have infrastructure that bad.

This historical situation provided researchers with a quasi-natural experiment. The researchers did not randomly assign iPhones to different regions, but AT&T’s preexisting telecommunications infrastructure created different degrees of access to the iPhone. They could therefore compare counties with better AT&T coverage against counties with worse coverage and observe whether the two groups experienced different changes in birth rates after the iPhone went on sale.

The study found that fertility fell earlier and by more in regions with better AT&T mobile-broadband coverage. It estimated that the introduction of the iPhone reduced births among women aged 15 to 19 by 4.5 to 8 percent, and births among women aged 20 to 24 by 3.2 to 6.6 percent. Smaller declines also appeared among older women.

Combining the estimates across different age groups, the researchers proposed a fairly striking figure: The spread of smartphones may explain 33 to 52 percent of the decline in the US general fertility rate among women aged 15 to 44 since 2007. If this result holds, smartphones are no longer merely one minor factor among many. They may explain one-third, or even close to one-half, of the United States’ most recent fertility decline.

The researchers also used two other equally lousy American telecommunications companies, Verizon and Sprint, to conduct placebo tests based on those companies’ network coverage before 2011. At the time, neither company sold the iPhone. If the real factor was simply better mobile-network coverage, greater regional prosperity, or more advanced communications infrastructure, areas with better Verizon and Sprint coverage should have shown similar results. The study did not find the same effect, which strengthens the credibility of the iPhone explanation.

The researchers also used the nationwide American Time Use Survey and surveys of sexual behavior to look for possible mechanisms. The data show that, as smartphones spread, people spent less face-to-face time with friends, had sex less frequently, and consumed more pornography. These changes are consistent with a mechanism in which smartphones reduce birth rates by reducing in-person interaction and partnered sexual activity.

The US study nevertheless has several limitations.

First, AT&T’s network coverage was not randomly assigned. Areas with better coverage may have had higher population density, stronger economic activity, and younger populations, and they may also have adopted other digital technologies earlier. Researchers can statistically control for some of these differences, but it is difficult to guarantee that every relevant factor has been excluded.

Second, the study measures changes in birth rates and may not fully distinguish between births that were forgone and births that were postponed. A 24-year-old woman who does not have a child because of smartphones, changes in social life, or some other reason may make up for it when she is 34. In that case, the short-term birth rate would decline, but lifetime fertility would not necessarily fall by the same amount.

Postponement cannot be completely ignored, however. Female fertility is constrained by age. If the first stable relationship is delayed and the age at first birth rises, the window for having a second and third child narrows. Some postponement may therefore ultimately become a permanent reduction.

Third, the iPhone’s effect may contain multiple mechanisms. Early smartphones provided internet video, games, social media, pornography, and dating apps. The study has difficulty distinguishing which function played the main role. It is therefore more accurate to describe the “smartphone effect” as the effect of the overall expansion of mobile digital life.

Fourth, the decline in teenage fertility has obvious positive significance in its own right. Few people would regard reducing underage pregnancy as a failure of public policy. The study’s finding that smartphones reduced fertility among girls aged 15 to 19 does not mean society should pursue a higher teenage-pregnancy rate. What truly deserves attention is that the change also extends to people in their twenties and thirties, as well as other age groups, and may affect the family lives people ultimately hope to attain.

This study offers evidence stronger than an ordinary correlation, but it does not deliver definitive proof of causation. Its safest conclusion is that the early spread of the iPhone across different parts of the United States was consistently associated with an additional decline in local fertility. Multiple tests and behavioral data are consistent with a causal mechanism, but the precise magnitude of the effect remains uncertain.

A common global technology shock

The second study covered by The New York Times is a global study conducted by Nathan Hudson and Herman Moscoso-Bodo of the University of Cincinnati. They examined data on smartphone adoption and fertility rates in 120 countries.

Global research faces a basic difficulty. High-income countries generally have relatively comprehensive telecommunications and time-use data, allowing researchers to estimate roughly when smartphones became widespread among teenagers. Data from countries such as Morocco, Indonesia, and Senegal are less complete, so researchers need proxy indicators.

Hudson and Moscoso used the point at which mobile-phone subscriptions reached a certain penetration level to estimate when digital mobile technology entered mass daily life. John Burn-Murdoch’s Financial Times analysis used another proxy: the frequency with which users in different countries searched for Android app stores. Rising search volumes for Google Play and other app stores can roughly reflect when smartphones began to be widely used.

None of these indicators is perfect. Having a mobile-phone subscription is not the same as having a modern smartphone, and app-store searches reflect only some user behavior. But whether 2012, 2013, or 2014 is used as a country’s technological turning point, the downward inflection point in fertility generally appears around the same time.

The most widely discussed chart included Australia, the United States, Britain, Indonesia, Mexico, Senegal, Egypt, and Iran. These countries differ enormously on almost every level of social organization. Yet after smartphones entered mass daily life, their fertility curves took on remarkably similar shapes.

Researchers summarize this phenomenon as a common global technology shock. Their logic is that if falling fertility in the United States arose from the financial crisis, the decline in the Nordic countries arose from changes to welfare systems, the decline in the Middle East arose from political, religious, and family-structure changes, and the decline in Latin America arose from economic fluctuations, then we would need a large number of country-specific explanations. Digital technology offers a common variable acting on all these societies at the same time.

The advantage of cross-national comparison is its scope. Australia was affected far less severely by the financial crisis than the United States and Western Europe, yet it experienced a similar fertility decline. Some economies in the Middle East and Southeast Asia continued to grow while fertility still fell. We can therefore weaken the claim that the 2008 financial crisis explains everything.

The limitations of cross-national research are equally obvious. Smartphone adoption may have occurred simultaneously with women’s education, urbanization, contraceptive technology, the global spread of culture, the development of streaming media, and rising incomes. Researchers have difficulty precisely separating these factors across 128 countries with data of varying quality.

Global research can therefore support three main conclusions. Between approximately 2007 and 2015, many countries experienced a new downward inflection point in their fertility rates. These turning points occurred very close to the time when mobile internet and smartphones entered mass daily life. And the financial crisis, income, housing, or cultural changes within any single country cannot fully explain this global synchronization.

As for exactly what share of the decline was caused by smartphones, we still lack a precise figure that can be applied universally across countries.

After the Financial Times report appeared, Burn-Murdoch gave numerous interviews. In one recent interview, he acknowledged that putting smartphones in the headline could create an overly simplistic impression. In his view, the real change involved the entire digital-media environment.

The internet existed long before the iPhone. The crucial change after 2007 was that the internet went from being a specific activity conducted on a home computer to a personal living environment that accompanied people anytime and anywhere. Social media, high-definition video, streaming media, video games, pornography, instant messaging, algorithmic recommendations, and short videos can all continuously occupy people’s attention through a screen they carry with them.

A more accurate research question would therefore be: How does mobile digital media change young people’s allocation of time, modes of socialization, and intimate relationships? The phone is the foundational device for this change, but what content people ultimately consume and how it affects them depends on the broader digital ecosystem.

If we truly want to turn the claim that smartphones reduce fertility into a credible explanation, we need to establish a stronger intermediate mechanism. Phones themselves do not suppress ovarian function, nor do they directly change sperm counts. What they influence is a chain of social processes that occur before reproduction.

That process runs roughly as follows: face-to-face socializing, meeting potential partners, forming romantic relationships, maintaining stable relationships, having sex, beginning to cohabit or marrying, and only then having children. A decline in the probability of any one of these stages will accumulate and ultimately be reflected in the number of births.

The loss of unstructured social life

The American Time Use Survey provides a set of very intuitive figures. Between 2003 and 2006, teenagers aged 15 to 17 spent about 12 hours per week outside school and work socializing face-to-face with friends. Today, that figure has fallen to less than six hours. It has been cut in half in only twenty years.

Other age groups have also experienced declines. People in their twenties, thirties, and even older age groups spend substantially less time socializing face-to-face with friends than they once did. The data from South Korea are equally striking. Since 2006, face-to-face socializing among people in their twenties has fallen by approximately half. Several European countries and Britain show similar trends.

Time is a finite resource. If someone spends an additional two or three hours each day using a phone, other activities are inevitably crowded out. Digital socializing can sometimes supplement real-world relationships, for example by helping friends stay in touch or helping marginalized groups find peers. But heavy use can also replace real-world interaction.

Sending messages, watching other people’s lives, and participating in online discussions can create a certain feeling of social connection, but they do not require people to share the same physical space or take on the risks of real-world interaction.

Researchers at the University of Cincinnati make a particular distinction between structured and unstructured time. Structured time is arranged by schools, families, institutions, or adults. Teenagers still have to attend school, participate in sports training, go to tutoring classes and clubs, and take part in other activities with explicit goals. Time spent on these activities has not declined by the same amount.

Unstructured time lacks clear tasks, evaluation criteria, and continuous adult supervision. Young people hang around after school, spend time at friends’ homes, or meet in neighborhoods, parks, shopping malls, or at parties. Their activities may have no specific purpose, and their plans may change at the last minute.

Adults often regard this kind of time as idleness or even as a source of risk. Young people may drink, get into fights, break rules, or engage in unwanted sexual behavior and experience unintended pregnancies. When teenagers spend more time at home using their phones, some social indicators really do improve. Teenage pregnancy declines, street violence and drunken altercations become less common, and some property crime also falls.

Yet unstructured time serves another function that is difficult to quantify: It gives young people practice in socialization. In those settings, people learn how to start a conversation, assess other people’s attitudes, deal with silence and awkwardness, express attraction, endure rejection, resolve conflict, and get along with their peers without an authority figure making the rules for them. Many of the skills required for romantic and intimate relationships gradually develop through these loosely organized interactions.

Structured activities cannot fully replace this kind of learning because, in classrooms, training sessions, or tutoring classes, everyone is acting around a shared task. In an intimate relationship, people must decide for themselves what to say, how to approach someone, how to respond, and whether to continue taking responsibility for the relationship.

Some people may believe that reductions in teenage dating and sexual activity are positive, and that there is no harm in waiting until adulthood to begin. The problem is that social skills do not automatically appear on a person’s eighteenth birthday.

Jonathan Haidt, the sociologist we often discuss on this podcast, emphasizes in his discussion of the phone-based childhood that early adolescence is a critical stage in the formation of personality, risk judgment, and social skills. A person who spends the years from 13 to 17 largely alone in front of a screen is unlikely to become, immediately upon turning 18, a fully socialized adult who is adept at forming intimate relationships.

The COVID pandemic gave the whole world an extreme example. Lockdowns and remote education forced young people to leave real-world social environments for a relatively long time. After the pandemic ended, schools and families around the world observed increases in teenage anxiety, avoidance, and social difficulties. The effects of digital media are not as concentrated as pandemic lockdowns were, but they may produce similar outcomes in a milder and more sustained way.

The decline in teenage pregnancy is therefore welcome in itself. A better path would be to reduce unintended pregnancy through sex education and effective contraception while preserving the real-world interaction, exploration, and social learning young people need. Moving all unstructured interaction onto screens may reduce immediate risks while creating a long-term deficit in relationship skills.

Before smartphones, a young person who stayed at home for long periods would become bored. Boredom exerts a kind of social pressure: It forces people to leave private spaces and look for friends and activities. Real-world socializing carries costs. It requires you to schedule time, leave the house, confront uncertainty, and possibly face rejection and failure. In an environment with few alternatives, people are willing to accept those costs.

Smartphones have greatly reduced the pain of being alone. Short videos, mobile phones, streaming media, social media, and pornography can provide continuous stimulation. Without coordinating schedules with anyone else, grooming one’s appearance, traveling to a public place, or risking rejection, a person can keep their attention satisfied.

Digital technology has changed the relative cost of real-world socializing. Real relationships can still provide deeper satisfaction, but they require more patience, coordination, and vulnerability. Digital entertainment provides shallow satisfaction, but it is faster, more reliable, and more controllable. When people are tired, anxious, and lacking in confidence, they often choose the option with less friction.

More choices, fewer relationships

There is a seemingly obvious objection: Social media and dating apps should help people meet more potential partners, so why might they instead reduce relationships?

Digital platforms really do expand the theoretical range of choices. A person can encounter people they could never have met in the past. The internet can play an especially positive role for people in geographically remote areas, sexual minorities, or those with unusual interests.

But expanding the range of choices does not automatically increase the rate at which stable relationships form. Having too many choices changes how people evaluate potential partners. Every actual person is placed within a seemingly endless list of candidates. As soon as friction appears in a relationship, the platform reminds users that vast numbers of alternatives remain available. People find it easier to continue searching, and they are also more likely to reduce complex human beings to a handful of comparable metrics.

Social media can distort people’s expectations of partners and lifestyles. Users see curated displays of appearance, consumption, travel, and emotion. Such content may raise standards of comparison, magnify the shortcomings of one’s own life, and make ordinary relationships appear less attractive.

Online platforms also often amplify hostility between the sexes. Extreme cases, stories of failure, and conflict-oriented opinions spread more easily. Young men and young women may enter different information environments and develop increasingly negative overall judgments of the other sex. Prejudices that might once have been corrected through concrete, real-world interactions are repeatedly confirmed in algorithmic environments.

Stanford University professor Alice Evans has pointed out that the stronger traditional gender roles are in a society, the greater the effect of smartphones on fertility. In such societies, women are exposed through the internet to new ways of life and new ideas about gender, reducing the appeal of traditional marriage. But men’s and women’s expectations regarding the division of labor within the family do not necessarily change in sync.

As a result, the two sides’ understandings of the terms of marriage gradually diverge, making partner matching more difficult. This also helps explain why fertility has fallen especially rapidly in some Latin American and Middle Eastern countries in recent years. Cultural changes brought by digital technology can enter people’s personal lives very quickly, while family institutions and gendered divisions of labor generally adjust much more slowly.

These reports also mention a factor that appears less frequently in discussions of fertility policy: online pornography. Smartphones have made pornography cheap, private, and available everywhere. A large body of research has found correlations between heavy pornography use and sexual dysfunction, reduced partnered sexual activity, and lower relationship satisfaction.

The direction of causation remains very complicated. Relationship difficulties may lead to greater pornography consumption, while pornography consumption may further weaken real-world intimacy. From the perspective of time use and incentive structures, pornography may serve as a substitute for partnered sexual activity among some groups.

Finding a partner, building trust, and maintaining a relationship all require long-term investment. Digital content can provide immediate sexual stimulation at extremely low cost. The long-term consequences may include reduced motivation to enter relationships, changes in sexual expectations, or greater difficulty engaging in real-world intimacy.

Discussions of low fertility often offer a simple explanation: Young people have changed their values and no longer want children, so falling fertility is merely the result of free choice. The survey data in the New York Times and Financial Times reports present a more complicated picture.

In many countries, when young men and women are asked to describe their ideal lives, they still choose a stable partner and approximately two children. Even in South Korea, where the total fertility rate is extremely low, the number of children people say they ideally want is often higher than the number they ultimately have.

There is a fertility gap: a distance between people’s ideal family size and the number of children they actually have. Part of that gap comes from economic costs. Part comes from being unable to find a suitable partner, from unstable relationships, or from relationships beginning too late.

Current low fertility therefore cannot be attributed entirely to autonomous choices not to have children. Some people explicitly choose a childfree life, and that decision should be respected. But a substantial number of others originally wanted children and were unable to complete the preliminary stages leading to parenthood.

From this perspective, the public significance of the smartphone hypothesis is not that everyone should be required to increase the national population. It is that we should examine whether a technological environment is preventing individuals from attaining the lives they originally hoped to have.

Why fertility subsidies have limited effects

If fertility is falling mainly because raising children is too expensive, then government efforts to reduce costs should have an obvious effect. Over the past twenty years, many countries have designed policies according to precisely this logic. Policy tools have included cash subsidies, child tax credits, parental leave, free or low-cost childcare, housing incentives, birth bonuses, and tax reductions for families with multiple children.

Some of these policies have improved families’ welfare and reduced the burden on parents, especially mothers. They have independent value. The problem is that their effect on national total fertility rates is generally quite limited.

Hungary provides a classic case. Viktor Orbán’s government has long treated increasing fertility as a major national policy objective. Hungary offers families preferential housing loans, childbearing subsidies, and large-scale tax reductions. Women who have three or more children can receive extremely generous income-tax treatment.

These policies have required enormous investment and did produce some recovery in Hungary’s fertility rate, but they did not restore the country to population-replacement level. The policy effect later began to level off as well. Even when economic incentives are increased to a very substantial level, very few people will create an intimate relationship that did not previously exist just because they can get an income-tax reduction.

Traditional policies operate on the latter half of the family-formation process. We can simplify that process into several stages. First, young people enter real-world social environments. Second, they meet potential partners in those environments. Third, they form and maintain stable relationships. Fourth, they acquire housing or economic security. Fifth, they decide to have children. Sixth, they bear the costs of pregnancy, childbirth, childcare, and education.

Traditional fertility policies mainly operate on stages four through six. Housing support improves family stability, while parental leave and childcare services reduce the costs incurred after a child is born. Cash subsidies affect couples’ decisions about whether to have a first, second, or third child.

If growing numbers of people remain stuck at stages one through three, these policies cannot reach the problem. Someone who rarely participates in offline social activity and has gone without a partner for years will not immediately enter a marriage because the government increases child subsidies. That person does not know how to enter a marriage. Increasing the number of nurseries also has no direct effect on people who have not yet formed relationships.

Declining rates of partner formation can therefore explain a policy paradox: Why does fertility continue to fall even as Western governments invest more and more in families? These policies have reduced pressure on people who already have children, but they have not created enough families.

Over recent decades, parents and institutions have constantly tried to eliminate risks from teenagers’ lives. Children’s schedules are filled with school, tutoring, sports, and all kinds of training. Free, unsupervised time is easily squeezed out by concerns about safety and traffic, as well as competitive pressure. Smartphones entered precisely this gap, providing young people with an alternative that appeared safe, quiet, and easy to supervise.

This arrangement reduced short-term risks, but it also weakened young people’s independence. Society needs to discuss how much disorder, risk-taking, and failure it can accept. Forming an intimate relationship inherently involves disorder, risk, failure, and uncertainty. A person must allow someone else to approach them while accepting that the other person may reject, misunderstand, or hurt them. An upbringing that excessively pursues safety and control makes it harder for young people to take on those risks.

This does not mean we should restore high teenage-pregnancy rates, drunken altercations, or violence. The goal should be to reduce serious harm through contraception, sex education, and reasonable protections while preserving opportunities for social exploration. Safety and growth require a more nuanced balance.

From screen time to shared life

I only recently discussed Jonathan Haidt on this podcast, along with legislation being considered by governments around the world to restrict teenagers’ access to social media. In the Financial Times report, Burn-Murdoch was cautious about governments directly reversing these trends.

He believes smartphones have become deeply embedded in education, work, payments, transportation, and social life. A comprehensive ban would be both unrealistic and enormously costly to enforce. He thinks change is more likely to come from social norms.

Growing numbers of Westerners have begun actively controlling their screen time, using basic phones that can only make calls and send messages, putting devices away collectively during gatherings, or setting their own limits on social-media use. Some Western schools and families are also developing a new consensus around delaying the age at which children receive smartphones.

The declining quality of digital platforms themselves may further encourage this shift. Early social media brought together friends, news, sports, culture, and political discussion, making it difficult for users to leave. Today, platforms are gradually fragmenting. Recommendation feeds are full of AI-generated content, engagement bait, and repetitive information. As the user experience deteriorates, the cost of leaving these platforms may decline as well.

Data in some reports already suggest that average global time spent on social media may have peaked several years ago. It remains uncertain whether this trend will continue. But changes in behavior do not always depend entirely on self-discipline. When a product is no longer interesting, leaving it becomes much easier.

At the end of the Financial Times report, there was also a discussion of a recent anti-technology wave among young Westerners. They are using basic phones, returning to offline life, spending time in nature, and regaining physical vitality through exercise. These changes may reduce screen use and improve health, but Burn-Murdoch raised a noteworthy reservation.

Today’s fitness culture is often combined with a highly individualized form of self-optimization. People calculate their physique, diet, sleep, step counts, and workout performance, turning the body into a project of endlessly improving performance. That way of life can still be highly self-centered.

A person can go to the gym every day without forming a deep bond with anyone. They can have a very strong body and an extremely disciplined routine while remaining unwilling to accept the disorder, compromise, and responsibility that come with a partner.

Returning from digital life to embodied life is therefore not the same as returning from isolation to communal life. Restoring Western fertility does not require more health projects completed alone. It requires more sustained, concrete relationships between people, relationships that cannot be completely controlled.

So, are smartphones causing a cliff-like plunge in global fertility rates?

The existing evidence is insufficient to support a definitive, single-cause conclusion. Global fertility has been falling for two centuries, and smartphones can explain only the most recent phase. Changes since 2007 have also been shaped by housing, economic insecurity, gender relations, years spent in education, the institution of marriage, and confidence in the future.

The two new studies have advanced the discussion. The US study used AT&T’s exclusive sale of the iPhone to turn variations in network coverage across regions into a quasi-natural experiment. The global study looked across 120 countries for a temporal relationship between smartphone adoption and turning points in fertility rates. The evidence they offer is stronger than a simple overlap between curves, and it also reveals a mechanism grounded in real life.

That mechanism begins with how time is allocated. As smartphones and digital entertainment increase, face-to-face socializing declines. Unstructured interaction declines. The opportunities young people have to learn social skills, explore relationships, and meet potential partners decline as well. Dating and sexual activity are postponed, stable partnerships become less common, marriage and cohabitation decline, and the ultimate result is fewer births.

No single stage determines an individual’s fate. But when small changes happen simultaneously to hundreds of millions of young people, they can alter global demographic trends.

These two Western reports also help us understand a fact that was easily overlooked in the past: Many young Westerners still want partners and children. Not everyone who ultimately remains childless made a firm choice to have no children. Some people never met a suitable partner. Some missed their reproductive window before a relationship could form. Some have spent years living in environments that are unfavorable to intimacy. The gap between individual ideals and actual outcomes constitutes an important part of contemporary low fertility worldwide.

The fertility issue therefore cannot be approached solely in terms of the costs incurred after children are born. Cash subsidies, tax reductions, parental leave, and childcare services can help families that have already formed, but they have difficulty addressing the decline in social interaction that occurs before family formation.

Effective policies still need to improve housing, employment, and gender equality. They must also seriously confront teenagers’ phone use, access to public spaces and unstructured time, and capacity for real-world interaction.

In the past, people worried that the global population would grow without limit and that resources would be unable to bear the strain. Today, growing numbers of countries face the opposite prospect: population aging, a shrinking labor force, rising fiscal pressure, and many people entering middle and old age without having fulfilled their own hopes for family life.

How much responsibility smartphones bear for this change still requires more time and more research to establish. But the concentrated coverage of the issue by the Financial Times and The New York Times over the past two months already marks an important shift in Western demographic debate.

Low fertility is no longer understood only as a result of the rising cost of raising children. It is also beginning to be understood as a decline in partner relationships. The decline is no longer only about how many children people are willing to have. It is also about how many people are able to enter a relationship capable of leading to parenthood.

Digital technology has not demanded that people abandon families. But it has created a new set of living conditions in which a person can receive continuous entertainment and stimulation without a partner, without friends, and while rarely leaving private space. Loneliness has therefore become easier to sustain, while the necessity of real-world interaction has declined.

Intimate relationships require time, physical presence, patience, and compromise. They require people to accept the risk of rejection and loss of control. Childbearing occurs after this entire sequence of relationships. If a society continually reduces people’s opportunities to meet, spend time together, learn, and build a shared life, then the number of births worldwide will certainly continue to decline.

The global fertility crisis is therefore simultaneously a crisis of social interaction, a crisis of intimate relationships, and a crisis concerning how human beings live together. The phone is only one technology involved, but it reveals the most profound change of this era: People still need one another, yet everyday life increasingly demands less and less that we truly face one another.