America’s Fertility Crisis Has Astonishing New Link to iPhones, Study Finds

iPhones may have slightly exacerbated an already underway drop in unintended pregnancies among teens. That's the big finding in a new working paper published by the National Bureau of Economic Research (NBER).

In a more sane environment, this paper would be greeted with somewhere between slightly increased respect for the poor smartphone and a collective shrug. Or perhaps with some skepticism—how exactly did the authors reach this conclusion anyway? Does it hold up?

You are reading Sex & Tech , from Elizabeth Nolan Brown. Get more of Elizabeth's sex, tech, bodily autonomy, law, and online culture coverage.

Alas, we live in a period of total paranoia and doom about smartphones. So the NBER paper is being heralded as a sign that smartphones are to blame for birth rates falling generally and a big, tragic harbinger of population doom.

Today, I want to look ( once again ) at why this fatalistic view is unwarranted and how the hype about phones and fertility doesn't hold up.

Should We Really Mourn a Drop in Unintended Births to Teens?

Let's start with the study itself. Authors Caitlin K. Myers and Ezekiel Hooper attempt to look at the iPhone's effect on childbearing by assessing U.S. fertility rates in places where AT&T provided mobile broadband coverage between 2003 and 2011 and places where it didn't. The iPhone was available only on AT&T networks during the period between June 2007 and February 2011.

From this, they conclude that iPhones did, indeed, lead to birth rates dropping. But even taking their calculations and explanations at face value, we're mainly looking at a phenomenon involving teen girls and, to a lesser extent, women in their early 20s. The results suggest "the fertility drop is concentrated among young populations and largely operates through declines in unintended births," the authors write, adding that one of the methods of suppression may have involved greater access to information about birth control.

Per Myers' and Hooper's calculations, "access to the iPhone reduced births by 4.5–8.0% at ages 15–19 and 3.2–6.6% at ages 20–24." But among 25- to 29-year-olds, the reduction was just between 1 percent and 1.3 percent; among 30- to 34-year-olds, it was as little as 2.7 percent; and among 35- to 39-year-olds, it was just 1.4 percent.

Is alarm really quite the right response here? Because even if we accept the underlying premises and conclusions of the study—and that's a big if , as we'll get to in a minute—what we're looking at here seems to be more girls and women avoiding unintended pregnancy at young ages or choosing to wait until they're more emotionally, financially, or professionally ready to have kids. If the iPhone really did depress fertility in this way, I'm not convinced that's a bad thing.

But About That Study Design

I'm also not convinced that there's really an iPhone-to-fertility-drop pipeline at play here. Remember, this whole experiment is based on studying "counties with near-universal AT&T coverage to counties with little or none over 2003–2011," as the authors put it.

The authors didn't measure how many people in their study areas actually had iPhones or whether birth rates actually dropped more among iPhone users. They just measured overall birth rates in areas with more or less AT&T coverage.

There are likely many differences between places where A&T coverage was extremely high and those where it was extremely low. More remote or rural areas would have had less coverage, while densely populated urban and suburban areas would have had more coverage. The former tend to be places where people are poorer, more religious, more isolated (including from access to birth control), have different norms, and so on. And keep in mind these were also the years of the Great Recession, which could have hit people in big cities and in small towns quite differently.

In short, there are all sorts of reasons independent of phones why births might have continued more apace in places with low AT&T coverage.

The authors attempt to control for this by "reweight[ing] control counties to match treated counties on observable demographics." But even if you control for certain aspects—income and education, say—it's hard to control for differences in cultural attitudes, community norms, economic and psychological effects of the recession, political leanings, access to contraception, and everything else that sets these areas apart.

It becomes really unclear: Are we looking at iPhone effects, or just urban vs. rural fertility trends during the Great Recession?

Even the study authors admit that it may be the latter. "Given that [high coverage] counties are systematically more urban than control counties, any other forces causing urban fertility to decline relatively more than rural fertility over this period could generate the same pattern."

Some problems with this study:

1.) There was already a pretrend in evidence. So they're claiming smartphones caused a pattern already in evidence.…