Should I let my baby watch TV?

7 min read

It’s 2pm on a Sunday, and my three-month-old son is watching the Jets game.

We have friends over for the afternoon. We’ve spent some time in the kitchen, eating, chatting, fielding questions about first-time parent life. And now we’re all sitting on the couch with the football game on, and my son is transfixed.

We laugh. He loves it! Who knew he’d be such an NFL fan?

Then we look at each other. Doubt begins filtering in. My husband squints his eyes and purses his lips—we probably shouldn’t be letting him watch TV, right? I look at him, the baby, the TV, the baby again. Try to position my face between the baby’s unblinking eyes and the possibly-brain-melting-but-possibly-totally-fine TV rays. The baby remains slack-jaw, his eyes casually flickering up over my head, getting a better look at QB Zach Wilson. I don’t know, I say.

Later that evening, our friends have left and I’m thinking about it more—is it bad for young babies to watch TV? As is the sacred ritual of every new parent, I pull out my phone and seek Almighty Google. Let’s see what the experts have to say. But before I can begin filtering through the jumble of medical advice, SEO-boosting baby product blogs, and momfluencers, I stop.

Wait. I’m an “expert” on this. I’ve definitely spent more time researching kids and screen time than the makers of Baby Bjorn or some Victorian Era-Inspired Momfluencer (a real thing).

As a scientist, I like to think of myself as someone who makes decisions based on evidence. I try to gather the facts, the relevant statistics, the latest research on a topic before deciding how to behave. I like to run, for example, and my mother lives in fear that something horrible will happen to me while I am out for a jog. She once texted me an article about a runner who was knocked out by a flying deer (the deer had been hit by a car a few yards away). Be careful! she wrote.¹ While it is certainly possible that an airborne deer will crush me mid-stride, as far as I can tell, the data suggests that these incidents are incredibly rare. I have not changed my behavior. On the other hand, when I was in college, a study came out suggesting that a single indoor tanning session increased risk for developing skin cancer by 75%. I have not looked in the general direction of a tanning bed since.   

My mental model for these decisions was always very linear. Gather scientific evidence, make decision. The data itself will contain the answer. But the problem with becoming a so-called expert on a topic—particularly one as muddy as children and screens—is that you know the data, and lack of data, too well. You can no longer see the forest for the trees. You’ve spent a decade examining each individual tree, writing pages on its strengths and flaws, noting which trees are missing or decaying slightly, checking which trees to trust and which will be easily felled with the tiniest axe to their midsections. You’ve forgotten that what you’re stumbling through is, in fact, a forest at all.

You realize that the data can only get you so far. There’s also preference, practicality, individual characteristics. You might even have to consult science’s conniving foil: common sense.

As it turns out, I know what the data suggests about screen time for babies.

The average child in the U.S. under age 2 spends 45 minutes per day watching TV or videos. And yet the American Academy of Pediatrics takes a hard line, recommending no screen time at all for children under 18 months (except for video chatting). Perhaps this discrepancy between medical recommendation and actual behavior is what has driven heightened interest in this question. Or maybe other researchers, like me, just want to know how guilty they should feel about letting their own babies tune into The Great British Baking Show. As they say, research is me-search.

Either way, multiple meta-analyses—studies that take all the other studies on a topic and combine the results²—have come out in just the past year on young children’s screen time in relation to sleep, language skills, and various other health outcomes (e.g., obesity). Here’s what they tell us: on average, there is a small association between screen time and negative outcomes for young kids. This is especially true when screen time is replacing other important activities, like physical activity or sleep. In particular, excessive screen time in the evenings seems to be linked to a later bedtime and shorter sleep duration. One of these meta-analysis estimates that children under age 7 who engage in excessive screen time are nearly twice as likely to sleep less at night.

This sounds bad. But there are a few challenges to interpreting these results.

First, the majority of studies—particularly those focused on infants—are “cross-sectional,” which means that all the data was collected at the same point in time. With these types of studies, we cannot figure out directionality: was it the screen time that caused the sleep problems? Or did kids who already had more sleep problems just watch more TV? Or was it some other factor that led to both more screen time and less sleep? There are also a few longitudinal studies, which take a group of babies, ask their parents about their screen time, and then follow their health outcomes over time. These help somewhat, but they still cannot fully account for other factors that might influence both the screen time and the health problems. What we really need are randomized controlled trials (RCTs).³ These are few and far between.

Second, these studies tend to focus on excessive screen time. For example, the health indicators meta-analysis set benchmarks for two sets of analyses at 1 and 2 hours per day, which, for an infant, is different than, say, 30 minutes here and there. Third, when it comes to language skills, we also find some positive outcomes of screen time when it is of high quality (e.g., Sesame Street) and co-viewed (i.e., watched together) with the parent. These effects, not surprisingly, are mostly found in kids over the age of 2, and they clearly muddy the “all screen time is bad” sentiment. And finally, many studies lump together kids across a range of ages, such as 0-18 months, or 12-24 months. This makes it difficult to tease apart the effect of a certain type or amount of screen time on a child who is 14 weeks and 3 days old (hypothetically speaking, of course).

In the end, this data leaves us with questions. What is the difference, in terms of risk, between TV exposure for a 3-month-old versus a 13-month-old? The NFL versus fast-paced cartoons? Sound versus no sound? 15 minutes versus two hours? Daytime versus nighttime? Tablet versus TV? How do we weigh these risks against other decision-making factors, like preference, practicality, and common sense?

Sometimes the path through the forest is not so direct.

Next weekend, the Jets will play again. My son will be one week older. And somewhere in the world, a scientist will be finishing up the latest study on children and screen time.

We will re-evaluate then.  

In the meantime, we’ll avoid plopping him down in front of a loud TV for hours-long stretches. We’ll try to limit his exposure to screens at night. We’ll make sure he’s getting plenty of sleep, and physical activity,⁴ and time interacting with adults.

And he’ll continue to sit on our laps, stealing glances at the TV, while we stare at him, wondering where on earth we could find an expert to ask about this.

Thoughts on this post? Ideas for future topics? Reply directly to this email!

1. Admittedly, she was half-joking. She also pointed out that the best part of this article was how bitter this woman seems toward the deer. I thought that the run would make me feel better after work, she says. And after they cut off her shirt, which was covered in deer blood? That’s when I knew a deer was part of this. You can almost see her staring, dead-eyed, at the deer and shaking her head at the audacity.

2. Meta-analyses are one of the best tools we have for summarizing what the evidence suggests about a given topic, but they are an unbelievable amount of work. As an example, want to know how many articles the authors of the health indicators meta-analysis had to manually screen? 39,657. How about the total number of participants across all studies included in the language skills meta-analysis? 18,905. Thank you, meta-analysis authors, for everything you do.

3. It is hard to do randomized trials like these in human infants for ethical reasons – we can’t rightly assign some babies to lots of screen time and others to no screen time and just see what happens. To get around this, some researchers developed an experimental protocol in which mice were exposed to “excessive sensory stimulation,” including loud noises and bright lights—perhaps the mouse equivalent of cartoons—for 6 hours per day, 42 days in a row. Initial findings suggest that mice exposed to these conditions show poorer cognitive functioning. I’m not yet ready to change my behavior with a human infant based on these results, but they are interesting. If only to imagine tiny mice glued to a TV while their mouse parents fret about the effects of screen time.

4. Is that a thing for infants? Turns out, it is.