What If TV Isn’t Bad for Us?

BRADY BUNCH THEME: Here’s the story of a lovely lady, who was bringing up three very lovely girls.

FLINTSTONES THEME: Flintstones, meet the Flintstones. They’re a modern stone age family.

The television set as we know it today was invented by a 21-year-old man named Philo Taylor Farnsworth, who’d grown up in a log cabin without electricity. His first broadcast was a straight line, which went out on the airwaves in 1927. By the mid 1940s, only about 6,000 U.S. homes had a T.V. set. By 1954 — just a year after the introduction of color T.V.s — half of all American households had a set. At the start of the 21st century, the average American home had more than two T.V.s. And in 2022, some of those T.V.s look like artwork hanging on the wall.

And the more T.V. Americans watch, the more we worry about how it affects us. Over the decades, Congress has held countless hearings and inquiries on the effect of T.V., especially on kids. In 1995, it passed the Television Violence Report Card Act, which called for an assessment of the violence on broadcast and cable T.V. shows to be made public. And separately, hundreds of studies over the years have suggested that violent programming causes aggressive behavior.

But does it? What do we know about how television changes our behavior? Studying the impact that television has on our lives is incredibly difficult because it’s so hard to know whether T.V. is really the cause of the trouble. Even after hundreds of studies, researchers still have questions about exactly how, or even whether, T.V. changes us. When it comes to television and behavior, do we really know what we think we know?

From the Freakonomics Radio Network, this is Freakonomics, M.D. I’m Bapu Jena. I’m an economist but I’m also a medical doctor. Each episode, I dissect an interesting question at the sweet spot between health and economics. Today is all about television and movies: Does television change our behavior? Does it cause children to perform worse at school? Is there even a way to find that out? Do violent movies lead people to become more violent? And what about any beneficial effects of the small and big screen? Do they exist? Can they be measured?

Interestingly enough, some of the more important information we have about how T.V. affects viewers is actually thanks to an attempt to understand something else entirely — education. In the spring of 1966, a 41-year-old man named James Coleman waited in a dreary motel room in Washington, D.C. He’d been there for days, barely sleeping, just one change of clothes on hand, and a bottle of bourbon nearby. All around him were stacks and stacks of paper, and each day he waited for more to arrive.^,

Coleman was a sociologist at Johns Hopkins University, and he’d been charged with writing a report on the state of public education. The 1964 Civil Rights Act called for a survey on the lack of equal opportunities in education due to, quote, “race, color, religion or national origin.” The report had to be delivered to the President by July of 1966, two years after the Act was passed. And it was Coleman’s job to do it.

Coleman and his team surveyed nearly 600,000 students from first through twelfth grade, along with 66,000 teachers, from 4,000 schools across the country. The report he submitted to President Lyndon Johnson was more than 700 pages long and challenged long-held assumptions about what mattered most when it came to furthering equality in academic achievement. But the study, which became known as the Coleman Report, was also significant for the sheer volume of data it collected.

JESSE SHAPIRO: This was one of the most important pieces of empirical social science of the 20th century.

That’s Jesse Shapiro. He’s a professor of economics and business administration at Harvard. Several years ago, Jesse had questions of his own about education — namely, how did preschool T.V. time affect academics later on? And when it came time to design his study, he knew just what to reach for. Because all that data in the Coleman Report? It was collected from students who grew up during one of the 20th century’s most life-changing moments: the introduction of television.

Jesse and his colleague, Matt Gentzkow, an economist now at Stanford, thought the Coleman Report could give them a way to tease out whether or not television negatively affected academic achievement. Here’s why: Remember, the Coleman Report was based on a survey that began in 1964. Jesse and Matt focused on students who were in grades six and 12 at the time of the survey. So, these were children who’d been born between the late 1940s to mid 1950s. That time period was a crucial moment in the spread of television across the United States. The 600,000 children that Coleman surveyed for his study had some stark differences when it came to television watching.

SHAPIRO: Some 12th graders experienced the first few years of life in a place with no television. And some 12th graders experienced the first few years of life in a place where television was available the entire time.

Here’s the thing about the advent of T.V. — it didn’t spread across the country at the same time. Some places got it before others, due to a mix of different reasons. But one important one was an FCC freeze on the issue of new T.V. station licenses in certain cities. The key was this created a natural experiment where some kids were — really kind of at random — exposed to early childhood T.V. and others weren’t.

SHAPIRO: So, think first about two 12th graders, both born in 1948. One of them grows up in a place where television doesn’t arrive until, say, 1954. Now think about another 12th grader, also born in 1948, who grew up in a place where television licensing started before 1948. So, for that person’s entire preschool period, television is available. So, if, say for example, television viewing was really bad for cognitive development, one of the things that we might see is that the second 12th grader’s test score performance is worse than the first one’s.

So that was the thinking behind the study design. Because the arrival of T.V. across the country was random, we can eliminate factors like family involvement in test score differences. But the researchers did something else clever. They included one more comparison to make sure that geography wasn’t the explanation for higher or lower scores.

SHAPIRO: That’s where having the sixth graders is helpful, because the sixth graders in our data, they’re born in 1954. And so, regardless of where they’re growing up for the most part, they’re growing up in a place that has television available. So, by comparing the sixth graders between the two kinds of places, we can get a sense of what’s the baseline difference in test scores that’s due to the place itself. And then by comparing the 12th graders, we can look for the additional component over and above that, that seems to be due to the availability of television.

So, looking at test scores among the sixth graders who all grew up with Lassie and the Mickey Mouse Club, but lived in different locations, shows the effect of place — where a child grew up — on academic achievement. That makes it possible to account for the effect of geography when comparing the 12th graders who grew up in those same places that either did or didn’t have access to television when those kids were growing up.

The Coleman Report included data about these sixth and 12th graders’ television viewing in 1965, but didn’t have any information about how much time they had spent watching T.V. as toddlers. Jesse and Matt were actually able to find that data in a Senate report on T.V. and juvenile delinquency from the 1950s.

SHAPIRO: A lot of evidence pointed to television being very popular with kids right away. So, even in the very early years of television. We can say that the average kid in the survey who was in a place where television was available was probably watching a reasonable amount. And we also know from the early primitive forms of ratings something about the kinds of television that kids were watching. And there were a lot of shows that were targeted specifically at kids from very early on.

HOWDY DOODY: Hey kids, what time is it?

KIDS: Howdy Doody Time!

SHAPIRO: So, an example is Howdy Doody. But also, there were shows that were not really specifically targeted at kids that were popular among kids. So, I Love Lucy would be an example of something like that.

LUCILLE BALL: Hello friends, I’m your Vitameatavegamin girl. Are you tired, run-down, listless? Do you poop out at parties?

So, what did Jesse and Matt find? Did watching television as a preschooler lead to lower test scores later on?

SHAPIRO: If you look at the effect of having television as a preschooler and test scores, the estimates that we find are pretty small. So, sometimes they indicate a small, positive effect. Sometimes they indicate a small negative effect. which is a way of saying that we can use our data to statistically rule out moderate or large effects of television on test scores.

They looked at some other variables too, like how much time students spent on homework. The study, which was published in 2008 in the Quarterly Journal of Economics, didn’t uncover any long-term effects of toddler T.V. time there either. But there was one area where they spotted a difference.

SHAPIRO: We found that the effects of television are more positive for kids whose first language at home was not English. Television may have had some benefits in terms of language development, or at least English language acquisition. And that fits with other studies that try to look for whether kids learn words from television.

Jesse emphasizes that this study isn’t the final word on toddlers and T.V. Hundreds, if not more, studies have analyzed the impact of T.V. on academic performance and behavior. But many suffer from a fundamental empirical problem: it’s hard to separate the effects of T.V. from the environments kids are in. In other words, because kids aren’t randomized to more or less T.V. exposure, we can’t really say what effect T.V. has on kids. That’s why what Jesse and Matt did with the Coleman Report is so impressive — they were able to find a natural experiment in which, by chance, some kids were exposed to T.V. in early childhood while others weren’t. And they could track the kids’ outcomes over time.

It’s hard, though, to generalize their findings to kids and T.V. today. After all, Jesse and Matt’s study considered programs from the 1950s, not SpongeBob, Paw Patrol, or Ninjago. There was no Netflix. No Hulu. No Disney Plus. Today, there’s a much wider range of choices and therefore a lot more variation across the country in what shows kids watch.

SHAPIRO: The effect of television is different for different kids in different kinds of households. One possible explanation is it may be related to what kinds of activities television watching is replacing in the home. it would be really interesting to try to understand that better and see what kinds of activities is television or other screen time displacing? And how does that relate to the effects of television on things like test scores or measures of behavior in the classroom.

Whatever activities that kids aren’t doing while they watch T.V. — that’s what economists refer to as “the counterfactual.” And we’re going to come back to that later. But first, I want to turn to another interesting study about television.

OSTER: I have worked on a lot in my career issues around gender equality and how women are treated in in different areas and what kind of rights and freedoms that they have.

Emily Oster is a professor of economics at Brown University. You may have heard her before — she’s been a frequent guest on Freakonomics Radio, often to talk about parenting, a topic she’s become especially well known for during the pandemic. I also spoke to her and Steve Levitt on Freakonomics, MD, in episode 12, about what to do with our ideas that fail. Emily and Jesse Shapiro are actually married, but it’s just a coincidence that they both have studies on the effects of television. And Emily’s work comes at the question from quite a different angle.

Back in 2007, when she was a graduate student, Emily wanted to know if having access to cable television changed how women in India thought they should be treated by their husbands, families, and society. She and her supervisor at the time, Rob Jensen, an economist now at the Yale School of Management, used data from a study called the Survey of Aging in Rural India, or SARI for short. Every year from 2001 to 2003, researchers surveyed 2,700 households across India where there was at least one person over age 50. They asked all female residents aged 15 and older about their experience as a woman. Did they need to ask their husbands for permission to visit the market or go to a friend’s house? Were they allowed to have their own money to spend as they wished?

Domestic violence rates were high in India at the time, so the survey also asked women: Was it okay for a husband to beat his wife for being unfaithful or disrespectful? What about for leaving the house without his permission, neglecting the children, or if her cooking wasn’t good enough? It asked women whether they wanted their next child to be a boy or a girl, since there has traditionally been a strong cultural preference for boys. And there were some other questions about how free women felt to make decisions about their children and their own lives.

OSTER: One of those is pregnancy. So, looking at fertility rates. And then the second was school enrollment. So, looking at how much kids went to school as something that we know often to be linked with autonomy for women.

Emily and Rob took all the responses collected over the three years of the survey and matched it up with data on cable tv access.

OSTER: So, you’ve got one set of places that always have cable, one set of places that never have it, and then you have these intermediate places where they get cable either between year one and year two, or between year two and year three. And then we’re looking to see whether the attitudes change in those changer villages and whether the changes line up in terms of timing with the timing of the television introduction.

It’s important to mention that Emily’s study, which was published in 2009 in the Quarterly Journal of Economics, didn’t compare places that had cable T.V. to places that didn’t. Emily’s study relied on changes in cable access within the same city or village over time.

OSTER: They’re just really different kinds of places. One of them is, you know, metropolitan Delhi. And one of them is a rural village in the middle of Tamil Nadu, or Bihar, where — those are just totally different experiences. That’s why what’s most important is looking at what happens in the same village when it changes their access to T.V.

And when cable became available, suddenly there was a lot more to watch. There were some entertaining shows before cable, but most of the programming was either government-sponsored news or informational shows. But the new cable channels introduced game shows and soap operas that people just loved.

OSTER: The most popular soap opera, the title translates to “Because a Mother-in-Law was once a Daughter-in-Law Also,” and it’s about, you know, a family and somebody has amnesia and the mother doesn’t get along and then somebody comes back and he remembers who he was, but he’s fallen in love with the nurse. So, it’s kind of Days of our Lives, but in Hindi, basically. I think that’s what you want to imagine. Cricket and Days of our Lives in Hindi.

I used to visit family in India as a kid and thinking back, the stories on that show sort of broke with what was traditional in India at that time. And when Emily and Rob looked at how women responded to those survey questions about personal freedom before and after they started watching shows on cable T.V., they saw an amazing change.

OSTER: When cable T.V. is introduced, we see changes in some of these attitudes about spousal treatment, in the direction of going to be more women favorable. We actually also see some reductions in pregnancy rates, and we see some changes in school enrollment, particularly for girls.

And perhaps most notably, women were changing their views on domestic violence.

OSTER: There’s a list of five or six situations in which the question is just: “Is beating acceptable in this situation?” And so, in the pre-cable period people report an average of two circumstances in which beating is deemed acceptable. and after the cable television is introduced, that drops to something like 1.7. So, it’s basically a sizable, significant reduction in the number of situations in which this is reported acceptable. now there are some circumstances in which somebody in the sort of pre-cable period said that beating was acceptable, that they now think, “You know what? It’s not acceptable in that circumstance.”

As for the preference for a son, well, women were changing their minds about that too.

OSTER: Something like between 55 and 60 percent of people report that they would want their next child to be a son. And then there’s actually about a 10 or even more percentage point drop in that share after the introduction of cable. We interpret that result as something about the exposure that households have to this television, changing their perception of the relative value of boys and girls or changing their perception about, you know, what would it be if I — if I had a daughter? How would that be?

Considering these findings, it’s tempting to think about how shows could be designed to change harmful notions about gender or any other dangerous behavior. But Emily points out that efforts like this usually don’t work, and probably aren’t necessary.

OSTER: Everyone loves cable T.V., right? Cable T.V. is not broccoli. Cable T.V. is the greatest. And so, over time it’s just rolling out and rolling out and rolling out. And so, it’s not really something the government has to, like, step in and, and engage with. I think the reality is that most of the time when government and other actors try to make educational television, it isn’t very good. When I was a kid there was this show, ‘Saved by the Bell.’ And there was this episode where Jessie Spano got addicted to caffeine pills. that really made an impression on me. And I’m not sure that a government infomercial that was like, “Don’t take caffeine pills,” would have had the same effect, but I will say I’ve never taken caffeine pills after watching what happened to Jessie Spano.

Emily says she was surprised by the size of the change among women who got cable during the survey. She actually spoke about this study at greater length in an episode of Freakonomics Radio a while back called “Why Do People Keep Having Children?” Her findings show us that television can change us, exposing us to worlds and ideas that we might not otherwise encounter. Watching T.V. can shake up our thought patterns and open our minds to new possibilities for our lives.

But what about the dark side of T.V. and movies — all those fears about violent programming leading to violent behavior. After the break: Do violent movies make us more violent?

*          *          *

Hey there. A quick note before we dive into the second half of this episode. I love getting your feedback — what we’re doing well, things we’ve missed, ideas for new studies to dig into and questions to explore. Lately, I’ve been thinking, though, what a shame it is that no one else gets to see your ideas but me! So, what if you follow and share your feedback and thoughts with me on Twitter at DrBapuPod? That way, your fellow listeners can read your ideas and chime in. We can brainstorm together about the latest episode or the latest research that caught your eye! That’s DrBapuPod — D-R B-A-P-U P-O-D — on Twitter. I’m looking forward to it. And now, back to the show.

*          *          *

One day, Gordon Dahl, an economist at U.C. San Diego, and a colleague, Stefano DellaVigna, an economist at U.C. Berkeley, were chatting about how media influences violence. And they got kind of irked about it.

DAHL: There’s lots of studies that show a relationship between people who like to watch violent movies, committing violent acts, and that literature was pretty unsatisfying to us because it’s not obvious that the violent movies are causing people to be violent. It could just as well be that violent people like to watch violent movies.

It’s the problem we talked about at the beginning of the show. How do you know whether the relationship between what a person is watching and how a person is behaving is causal or just a correlation? They were convinced, though, that there was some relationship between seeing violence on the screen and acting violent.

DAHL: Laboratory experiments showed that if you take — it’s usually college students or children — and put them in a room and have them watch a violent media clip, they are more aggressive. They get neurons in their brains firing. They play more aggressively in little games. And so, we were pretty convinced by that literature because we felt like they were randomly showing some people a violent movie clip and others, not a violent movie clip. And so, we said, “I wonder if that aggression that we see in the lab in the short run, immediately after you see a violent movie clip, does it actually change behavior of people in the real world? Could what we saw in the laboratory be replicated out in the real world?”

The question for Gordon and Stefano was, how can you randomly expose some people to violent movies in a way that you could study outside of a laboratory.

DAHL: Then we realized that one of the advantages of blockbuster movies in the U.S. is that a lot of people watch them. And the other advantage is they happen a little bit randomly throughout the year. Most Februarys are filled with romantic comedies, but one February there was Hannibal, which is an incredibly violent movie. And what we said is, oh, when Hannibal came out on the opening weekend, about 12 million people went to see it. I wonder if there’s an accompanying spike in violence on the weekend that that violent movie is released and it turns out it’s not just Hannibal — some Julys, for example, will have a big blockbuster violent movie and other Julys will have more of a Disney film or something else, which is not as violent. We decided to take advantage of this naturally occurring variation. Let’s try to look at weekends when in one year there was a violent movie and another year there wasn’t a violent movie and see if crime goes up.

Gordon and Stefano relied on ratings of movie violence from a website called Kids in Mind. That site rates movies like Hannibal as strongly violent. Avenger movies, for example, are typically rated mildly violent. And a nonviolent movie would be something like a Disney movie or Runaway Bride — I love the romcoms by the way. Anyway, Gordon and Stefano collected revenue data to see how many people were seeing each movie on any given weekend throughout the U.S. They also gathered crime data from across the U.S.

DAHL: We collected data from police calls, where police show up at the scene and they write up a report. So, not just arrests, but you know, did they get called because there was an assault that was alleged. And we combine all of that data together to see if on the weekends, when there’s a big revenue burst for a violent movie versus a nonviolent movie, do we see crime go up?

They reported their results in a 2009 paper in the Quarterly Journal of Economics. What they found was not at all what they expected to find. First, they noticed that between six in the morning and six in the evening, there was no change in assaults or violent crime tied to the violent movies. That made sense to them. Not many people go to see matinees, so there wouldn’t be a big change in crime following daytime movie showings. But then they looked at the night data.

DAHL: Here’s where our first surprise came. We looked in the evening hours between 6:00 p.m. and midnight, when most movies are being played and most people are attending the movies. And what we found was not an increase in violence but a drop in violence. And it wasn’t a small effect. For every million people watching a strongly violent movie, we had a drop of about 1.3 percent in violent crimes. And if you saw a mildly violent movie — you know, those Avengers, Spiderman type movies — we saw a 1.1 percent drop in violent crimes. And on average, the violent movies as they get released, throughout different weekends, they result in about 52,000 fewer assaults throughout the year. So, about 1000 fewer assaults on a given weekend. So, that was surprising to us because we predicted that there would be more violent crime, not less.

So, what was going on? Why would violence be lower after violent movies?

DAHL: People, even going all the way back to Aristotle, thought that viewing violence would actually be cathartic. That was one of the arguments for the Roman gladiators. if you see people doing violent stuff, maybe that gets rid of their violent tendencies.

But Gordon and Stefano didn’t think this explained their observation. And they realized something: laboratory studies of movie violence typically looked at a random selection of people. But it’s not random in real life. The people who choose to see violent movies are the people who like these movies and may even be slightly more violent themselves.

DAHL: When you take a violent person and you put them in a movie theater, they’re pretty much incapacitated. It’s very difficult to commit a violent act while you’re in a movie theater, because it’s a very monitored environment.

So, Gordon is suggesting that those people with violent tendencies aren’t out committing crime between the hours of 6pm and midnight because that’s precisely when they are likely to be in a movie theater watching — rather than committing — acts of violence. So, what happens after everyone spills out of the theaters? Well, the data also revealed a drop in crime between midnight and six a.m., one that was even bigger than the one between six p.m. and midnight. Now, why would that happen? At this point, people aren’t at the movies, so you can’t say that they’re busy doing something else instead of causing trouble. But a lot of times, a night out at the movies means a break from the usual weekend pattern. On opening weekend of the latest Avengers movie, a lot of people aren’t doing what they would normally be doing on a Friday night.

DAHL: For many people, this might’ve been going to a bar or another place where violent acts are more likely to take place. And so consistent with this, especially in the early morning hours, there’s fewer cases of violent crime that involve alcohol. Not only that, it changes the whole evening’s activities, so that they’re less vulnerable to committing a violent crime.

And that change in the evening’s activities may be just as important, or even more important, than whatever impact the movie itself is having on viewers. The researchers know this because — that drop in violent crimes? It was bigger when the blockbuster movie that people were flocking to wasn’t violent.

DAHL: So, what I’m saying is when you see a violent movie, there is an arousal effect, but it’s dominated by the time use effect. If you take a movie like an Adam Sandler movie, which attracts young men, the people who commit violent acts or commit them more often than others — you get a much bigger reduction in violence because you’re not arousing them. You’re not increasing their aggression, but you’re taking them off the streets and changing their time use for the whole evening. And that’s how we conclude — it doesn’t mean that the lab experiments are wrong. It just means that the time-use effect dominates the arousal effect. It’s much bigger.

Of course, when it comes to Adam Sandler movies, I’m pretty sure Gordon is talking more Happy Gilmore, less Uncut Gems, which was — incredibly violent by the way. I’m still a little scarred by that one. But anyway, violent movies reduce violence because the people who might be violent are simply busy doing something else — they’re at the movies. But if you show people at risk for committing violent crimes a movie that they like but that isn’t violent, the crime rates drop even more. So, their findings don’t actually contradict the lab studies that show watching violence makes people more violent. We can tell that violent movies do trigger aggressive behavior since crime rates fall more dramatically after nonviolent movies compared to violent movies.

DAHL: Watching a violent movie is a safer place for a violent person to be than in a bar. But an even safer place would be to have them watching a movie that they love, which doesn’t have any aggression or violence in it. And then you’ll get the biggest bang for your buck. If you have a Disney movie that’s blockbuster, you don’t get the right people incapacitated or changing their activities. What we could do, though, is reduce the level of violence in movies, but still have them equally attractive to young men or other people who commit violence. It’s not movies, per se. It’s about time use. What if you could just get the same people to do midnight basketball?

There are plenty of unanswered questions when it comes to watching violence. Does it make us desensitized over time? What are the long-term effects, over a lifetime?

DAHL: I’ve never seen great papers that can say what’s the effect of growing up and watching violent media for 10 years on whether you’ll be desensitized and be a more violent individual, which I think is really what we’re worried about.

For as long as television and movies have existed, people have been fretting about the harm they might cause. And plenty of studies have validated those concerns. But the truth probably lies somewhere in the middle — violent T.V. shows and movies aren’t all good or all bad. They may not hurt academic performance, at least not in the way we usually measure it. They may expose people to ideas and lifestyles they’ve never personally experienced. They may make some people more violent. One thing we can say? The streets will probably be safer on opening night of the next Hannibal movie.

That’s it for Freakonomics, M.D. this week. I hope you enjoyed our discussion today and that you’re starting to think a little differently. You can find links to all the studies we mentioned at freakonomics.com — and also a very interesting episode of Freakonomics Radio that looks at another natural experiment — comparing young kids who did and did not have access to Sesame Street when it first came out. It’s called “Does Early Education Come Way Too Late?”

As always, I want to thank you for listening. It’d be great if you could give us a review on Apple Podcasts or wherever you’re listening. It helps new people discover the show. And like I said after the ad break, if you have any thoughts on the show, follow Freakonomics, M.D. on Twitter at DrBapuPod — that’s D-R B-A-P-U P-O-D — and tweet at me! You can also still shoot me an email at bapu@freakonomics.com.

*          *          *

Freakonomics, M.D. is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and People I (Mostly) Admire. All our shows are produced by Stitcher and Renbud Radio. You can find us on Twitter and Instagram at @drbapupod. This episode was produced by Jessica Wapner and mixed by Eleanor Osborne. Original music composed by Luis Guerra. The supervising producer was Tracey Samuelson. Our staff also includes Alison Craiglow, Greg Rippin, Mary Diduch, Rebecca Lee Douglas, Morgan Levey, Zack Lapinski, Ryan Kelley, Jasmin Klinger, Emma Tyrrell, Lyric Bowditch, Jacob Clemente and Stephen Dubner. If you like this show or any other show in the Freakonomics Radio Network, please recommend it to your family and friends. That’s the best way to support the podcasts you love. As always, thanks for listening.

OSTER: I watch a lot of The Real Housewives. The Great British Baking Show.

SHAPIRO: I Love Lucy. Saturday morning cartoons.

DAHL: Borgen. It’s kind of like West Wing, but in Denmark. The Crown. The Mandalorian. I think it would probably be classified as mildly violent. So, maybe you’re incapacitating me on some weekends.