What happened to American birth rates after 2007 is turning into a more complicated story.
A new working paper published by the National Bureau of Economic Research is making an argument that has not previously been quantified at this scale: the spread of the smartphones beginning in 2007 may have contributed meaningfully to the decline in U.S. birth rates in the years that followed.
The study was conducted by Caitlin Myers, an economics professor at Middlebury College, and her stepson Ezekiel Hooper, a 2025 Middlebury graduate. Their methodology focused on a natural experiment created by AT&T’s exclusive carrier agreement with Apple at the time of the iPhone’s launch. Because AT&T coverage varied significantly by county, the researchers were able to compare birth rate changes in areas with high iPhone availability against those with little or none between 2007 and 2011.
The results were notable. Among women in their 20s living in counties with extensive AT&T coverage, birth rates fell by 14.6% over that period. In counties with no AT&T coverage, the same age group saw a decline of 10%. Among teenagers, the gap was wider. Teen birth rates in counties with near-universal coverage dropped by 26%, compared to 13.8% in counties without it. The overall finding suggests the iPhone’s availability alone may account for a 33% to 52% decline in birth rates among women ages 15 to 44 during those years.
The paper has not undergone peer review.
Smartphone
Two additional working papers published through the Social Science Research Network in April and June of this year reached broadly similar conclusions. Both were authored by Hernan Moscoso Boedo, an economics professor at the University of Cincinnati, and Nathan Hudson, a doctoral candidate. Their research examined how digital technology broadly, not just smartphones, has affected social behavior and fertility patterns.
The more recent of the two studies found that 43% of the U.S. fertility decline since 2007 can be attributed to digital technology becoming cheaper and more widely available. The researchers argued that digital tools have fundamentally shifted how people allocate their time, favoring frequent but shallow online interactions over the sustained in-person contact that deeper relationships require. As those deeper relationships form less often and become less stable, the likelihood of having children decreases alongside them.
The core argument across all three papers is not that smartphones make people want fewer children. It is that they are replacing the unstructured, in-person social time through which partnerships and sexual relationships develop. Hudson pointed to the American Time Use Survey, which documented a 44% decline in in-person socializing among teenagers between 2003 and 2019, as supporting evidence for that mechanism.
What skeptics are saying
Not everyone finds the framing convincing. Sarah Hayford, director of the Institute for Population Research at Ohio State University, said she is open to the idea that smartphones have played some role in birth rate trends but questions whether a narrow five-year focus captures enough of the picture to be useful.
Her concern is methodological as much as substantive. The existing evidence on declining teen birth rates in the U.S. points strongly toward increased contraceptive use as the primary driver, not reduced sexual activity. If teenagers are simply spending more time online rather than in person, the mechanism the studies propose would predict less sex, but the data suggests contraception is the more significant factor. Those two explanations are not easily reconciled.
Hayford also placed the smartphone argument in a longer historical context. Research from the 1960s and 1970s found that the spread of radio and television, which depicted smaller family structures as normative, correlated with declining birth rates. The same logic may apply to parenting content on social media today. Falling fertility is a global phenomenon playing out across vastly different cultural and technological contexts, and she suggested that highly specific micro-level studies may not be the most productive lens for understanding it.
The three new papers add an interesting data point to a long-running debate. Whether they change the broader consensus is a question peer review has yet to answer.
Source: USATODAY