Does “Early Education” Come Way Too Late? (Replay)

STEPHEN DUBNER: Hey, Levitt.

STEVEN LEVITT: Dubner, how are you doing?

DUBNER: Great. Welcome to New York.

LEVITT: Yeah, it’s good to be here.

DUBNER: We don’t often get to do this, sit down face-to-face.

LEVITT: Yeah, because the traffic in New York is so terrible that I stay away.

DUBNER: Why are you here? You meeting with the Pope today?

LEVITT: No, the Pope passed on me.

DUBNER: Oh, you meeting with the U.N. today?

LEVITT: U.N. passed on me as well.

DUBNER: Just me?

LEVITT: Yeah, just you.

DUBNER: Alright. So, Levitt, I understand you’ve gotten into the early-education business?

LEVITT: Yeah, I think we regret it a little bit. Roland Fryer, John List and I went out and — really John more than us — started a pre-school, with the idea that if we were really smart and really thoughtful we could do things for three-, four- and five-year-olds that would change their lives and maybe could be used by other people to change a whole lot of lives all across the world.

Steve Levitt is my Freakonomics friend and co-author. He’s an economist at the University of Chicago. John List is a colleague there.

JOHN LIST: Stephen Dubner, how are you doing, my friend?

DUBNER: I’m doing great. Okay John, so tell us a bit about your work.

LIST: My primary research goes to the field and runs field experiments on issues like: Why do people give to charitable causes? Why do people discriminate against one another? Why do women earn less money than men?

List grew up in Wisconsin, went to college there and in Wyoming, then taught in Florida and a few other places.

LIST: When I moved to Chicago, I became very interested in public education largely because I saw the problems of public education all around me. I lived just south of Chicago in a city called Flossmoor. And right next to that is a city called Chicago Heights, and in Chicago Heights you have ninth graders who are reading at a fourth-grade level and doing math at a third-grade level. You have poor kids who just aren’t experiencing the value-added that children of the affluent experience. So, combining that with the fact that that’s sort of how I was raised myself — I was raised in public education and not in a wealthy family at all (my father’s a truck driver and my mother’s a secretary), so this was a sort of setting that I had been raised in — and I could see the problems in that setting, but they really came to fruition for me when I moved to the University of Chicago.

So List, Levitt, and the Harvard economist Roland Fryer set up an experimental pre-school in Chicago Heights.

LEVITT: The demographics are about one-third white, one-third Hispanic, one-third African-American. It’s a relatively poor community. And the idea there was to look at state-of-the-art techniques for teaching kids reading, writing, and arithmetic — cognitive skills — and compare the outcomes of kids with another curriculum that emphasized non-cognitive skills like sitting still and expanding working memory and executive function and things like that.

Even though the school was a school, it was also very much an experiment.

LIST: So, our overarching theme is: let’s go into schools and use them not only to teach our kids, but to teach ourselves what works and why.

But the school wasn’t the only part of their experiment.

LEVITT: The more interesting piece of what we did was to get away from the idea of the school and say, “Well, schools only have kids for a handful of hours per day, but who, really, will mold kids through their lives are the parents.” And so, if we could take the parents of pre-school kids and have an academy where we actually teach the parents how to teach their kids, then maybe we could have a bigger and a longer-term impact on the kids.

DUBNER: So you set up also then a Parent Academy, yeah?

LEVITT: Yeah, so we set up a Parent Academy.

LIST: And this was in a large basement room in a K-12 school in Chicago Heights. We have two teachers in there. The parents arrive for a session, which is once every two weeks for 90 minutes, and it’s over a nine-month-long period. And we then incented the parents for things like their attendance at our early childhood sessions. They were incented based on the homework assignments that their kids handed in. They were incented based on the interim assessments that we actually conducted with their kids.

LEVITT: So, over the course of the school year, parents can earn up to $7,000.

DUBNER: Wow. So, you’re trying to help kids do better at school by essentially paying their parents to teach them to do better at school, yes?

LEVITT: Exactly. To take parents who care enough about their kids that they’re willing to show up to these classes and equip them with a better set of skills so that they could be better teachers to their kids — not just, say, the kid that was in our program but all the siblings — and to do that over a lifetime as opposed to just one year.

DUBNER: And why go to the parents? Why not just pay the kids directly?

LEVITT: Well, for one thing, the kids were three-, four-and five-years old so they wouldn’t understand the value of money. Maybe trinkets could have done it. But I think we really believed that the power of success lies in the parents.

Levitt and List were pretty enthusiastic about their experiment – and undaunted. How hard could it be to set up a pre-school and a parent academy?

LIST: Okay, so now is where the real problem starts.

LEVITT: It was one of the most difficult things I think we could have tried to do.

LIST: So, we have to make sure that the superintendent, of course, is on board.

LEVITT: And as economists I think maybe we weren’t fully equipped to make it — it seemed like it’d be easy.

LIST: We needed to secure state licenses to run a pre-K program.

LEVITT: The logistics are hard.

LIST: We need to develop a partnership with the community so the community trusts us and they will actually send their kids to our Parent Academy.

LEVITT: It’s just plain complicated.

LIST: We need funding. Funding for the teachers, funding for development of the curriculum, and we needed funding for the incentives.

DUBNER: Wow. Okay, a couple quick questions before we get on to longer questions. What’s all this going to cost and where’s the money coming from?

LIST: So this particular experiment cost nearly $1 million and the money came from Anne and Ken Griffin.

DUBNER: Who have a foundation in Chicago, yes?

LIST: This was a foundation in Chicago. It is no longer a foundation in Chicago.

DUBNER: You put them out of business with your Parent Academy?

LIST: We put them out of business. We spent all of their money on these interventions and they’ve gone broke, so now Ken’s trying to make more money and so is Anne and maybe, maybe they’ll send us some more in the future.

The truth is, the foundation shut down when the Griffins divorced. But the economists did get their money to run the pre-school and parent academy. So, what did the Griffins’ foundation get for all that money? How did the experiment work out?

LEVITT: It worked out okay.

Just okay? Today on Freakonomics Radio, we’ll hear about their results. We’ll also hear about other early-education ideas, some of which are really cheap. And some early-education ideas suggest that even pre-school may be too late.

*     *     *

So, three economists, led by John List, decide to set up a pre-school in Chicago, with an experimental curriculum.

LIST: So, these are three-, four- and five-years olds, and our goal, of course, was to set up a program so we could measure how the children’s cognitive and non-cognitive — or executive-function — skills changed because of our program.

None of the economists knew much about this field.

LIST: So, essentially what we did is we started out doing a series of interviews of experts in early-childhood and development psychology. And we ended up hosting a two-day conference at the University of Chicago where we had the leading experts come in, and we talked about exactly what should the tools be that we use to measure cognitive and executive-function skills. And in the end, we settled on these assessment tools called Woodcock-Johnson on the cognitive side and on the non-cognitive side, things that measured operation span, spatial conflict and self-control.

The economists also set up a school for the parents of the pre-schoolers.

LIST: And here, we gave them a curriculum that was composed of both a cognitive component, which was called Literacy Express, and a non-cognitive component, which was called Tools of the Mind. So we combined these two curriculum, to try to push these two types of skills into the parents’, let’s say, minds, and then showed them how they could teach their children how to have these skills of self-control and to finish your homework.

And the parents, you will recall, were incentivized to do their work. A parent could earn up to $7,000 based on their own attendance and how well their kids did on homework and on assessment tests. So, what were the results of all this educational experimentation? What do the data have to say? Well, the research team is still working up the results on the pre-school curriculum. But as for the Parent Academy – John List, Steve Levitt, and Roland Fryer recently published a working paper on that project. The verdict?

LIST: Let’s jump right into the results.

LEVITT: These are huge, huge gains.

That is to say, incentivizing and teaching parents how to teach their kids really worked.

LIST: So, a first thing to note is we have huge differences across kids. Now, what I mean by that is that our program really, really helps Hispanic and white students and it doesn’t help blacks at all.

DUBNER: How disappointing was it to see that Hispanics and whites moved a lot and blacks didn’t?

LIST: Yeah, I think it’s a mixed bag. I think that when we started down this research agenda, part of our mission was, first of all, to learn about the racial achievement gap and learn about how we can lower that gap. Most of the time we lump Hispanic and African-American kids together, and we say, “What is the solution for minorities in public education?” And I think this particular result teaches us that the education production function across these two groups is very different and the solution will not be the same across African-American and Hispanic families.

DUBNER: Do you have any ideas for African-American families?

LIST: We have a few things boiling, but nothing that has popped up that moves African-American families.

LEVITT: And to be honest, we have no real theory for why that is. The two sets of parents were equally engaged in the program and we can control for all sorts of background characteristics and nothing really explains it. So to me that’s really a puzzle, and a puzzle that I don’t have an answer for.

But again, the Parent Academy did work really well for some students.

LIST: In just a nine-month program, we can move an Hispanic student from around the 30th percentile in test scores to above the 50th percentile in test scores.

LEVITT: These are bigger effects than almost any educational policy that is put into place.

LIST: So we’re getting results that swamp a result that you would get from a year-long program in a charter school. Now, that’s on both the cognitive assessments and the non-cognitive assessments.

But in looking carefully at cognitive skills versus non-cognitive skills, the researchers discovered something — something that was both a key to their success but also a warning sign.

LEVITT: Upon entry into the program, we tested kids to see where they stood in terms of their cognitive skills — how well they could, you know, do the alphabet and math and whatnot — and also their non-cognitive skills, about how well they could sit still and keep things in memory. And what was incredibly interesting to me in our findings is that for the kids who were below average on these non-cognitive skills — the ability to concentrate, to remember things, to kind of think their way through problems — the below-average kids made no progress in our program. So if you started behind, in terms of how ready you were to learn in some sense, then you got nothing out of our program. And that was true whether you scored high on the cognitive scores or not. So, kids could be really high achievers in terms of math and reading but gain nothing from our program if they didn’t have these sort of sit-still skills. But on the other hand if you were above average on these non-cognitive skills, you got huge benefits from our program. So what does this mean? Well, for one thing, I think it intuitively makes sense — that there’s a threshold for being able to learn. If the kids can’t concentrate, it’s hard for them to learn and no matter how hard the parents try it’s going to be hard to make gains. On the other hand, what it’s really valuable for from the perspective of public policy is that it really tells you where to target your resources.

So, what does the Parent Academy research suggest? Well, at least a couple of things. One: the fact that it worked — that at least some kids did better after their parents got tutoring – seems to confirm what Steve Levitt said earlier, that schools themselves may not have as much influence as we think.

LEVITT: Well, really schools only have kids for a handful of hours per day but who really will mold the kids through their lives are the parents.

Two: if a child’s ability to learn is so heavily influenced by what happens at home — and if the kids who came in with low non-cognitive scores at age three or four didn’t make cognitive gains — well, maybe the key here isn’t just what happens at home for a kid, but what happens at home well before a kid is even old enough to go to pre-school. In other words, what happens starting at age zero.

DANA SUSKIND: The brain, unlike any other organ that we have, comes out relatively underdeveloped.

 That’s Dana Suskind.

SUSKIND: Unlike your kidneys — they function from day one, as they will for their entire lives — the brain is underdeveloped, and it’s absolutely dependent on what it encounters on its ride to full development. And it’s really in those first three years of life when a huge amount of the physical brain is built — literally 80 to 85 percent of it. And what is building those connections — those 700 to 1,000 new neural connections — is language.

Suskind is one of those multi-hyphenates.

SUSKIND: I am the director of the Thirty Million Words Initiative and author of Thirty Million Words: Building a Child’s Brain. And I’m at the University of Chicago and a scientist, a surgeon, social activist.

DUBNER: Whew, exhausting already.

SUSKIND: And mother of three wonderful children.

Suskind is head of the Pediatric Cochlear Implantation Program at the University of Chicago, which means she sees a lot of families whose babies don’t hear well, if at all.

SUSKIND: So when they come to me, they’re often shocked, because most children born with profound hearing loss actually have no family history of it, even though it’s genetic. So first you sort of wrap your arms around this overwhelmed family, because the baby has no clue, and start sort of explaining what we can do.

What they can do has changed a lot in recent decades.

SUSKIND: It’s really a golden age for children with hearing loss. What they can expect out of life is what any child with typical hearing can.

And that’s because of what people like Suskind now routinely do.

SUSKIND: What I do each and every week is I implant this technology into the baby’s cochlea, which is sort of a snail-like-shaped structure where the nerve part of hearing begins. Then about two weeks later is really when their hearing birthday happens. You think it’s when you do the surgery; I have to tell parents, “Look, they’re not going to hear you right after surgery.”  But you know, two to three weeks later, the audiologist, with usually a whole gaggle of family members, are there to turn it on. 

DUBNER: So they really flip a switch at some point, yes?

SUSKIND: They literally flip a switch.

AUDIOLOGIST: Now, again, keep your eyes on her. And within a few seconds it will go on.

MOTHER: Natalie! Natalie, honey. Hi! Hi, baby!

AUDIOLOGIST: Awww. She’s so happy.

MOTHER: Do you hear mommy? You hear mommy.

But Suskind learned that giving kids the physiological ability to hear didn’t mean they’d automatically succeed in learning language.

SUSKIND: You know, how do we learn language? It’s not like all of a sudden when a baby’s born they come out speaking in sentences and understand what you’re saying. Hearing is not the ear; it’s the brain. And it takes time for the brain to understand what those sounds mean.

That’s what Suskind came to understand during follow-up visits with her patients.

SUSKIND: If I do a cochlear implant on a child at 18 months, versus five years of age, it’s a huge world of difference.

That is, the older ones had a much harder time learning language. And that’s because the brain’s neuroplasticity — its ability to develop with new stimuli — is most acute when children are under the age of three. Suskind found that age wasn’t the only factor working against some kids. Poverty was also a big issue.

SUSKIND: When you take the Hippocratic oath, your responsibility doesn’t end when the surgery ends, right? It ends when your patients do well. And for me, that meant understanding that there were so many, what we call, social determinants of health outside the operating room, which were impeding how my patients did.

So Suskind started a program called Project Aspire.

SUSKIND: Project Aspire is for children with hearing loss from low-income backgrounds. And as I was working on Project Aspire and learning, I realized, “Oh my gosh, what’s going on in my deaf population really just mirrors what’s going on in the typically developing population at large.”

And that led to Suskind founding another organization, called the Thirty Million Words Initiative, which gets its name from a remarkable study from the 1960s.

SUSKIND: Researchers Betty Hart and Todd Risley started off during the War on Poverty in Kansas City working with pre-school students, trying to sort of improve their vocabulary, trying to impact the achievement gap.

Hart and Risley, who were psychologists, studied the language environments of young children in 42 families.

SUSKIND: And they followed them for about two-and-a-half years ‘til the age of three, going into the homes every month, recording, trying to understand what the language environment looked like. And what they found, in a nutshell, was that by the end of the age of three, children born into poverty will have heard 30 million fewer words than their more affluent peers.

DUBNER: Wow, by the age of three?

SUSKIND: By the end of the age of three. So right before they were turning four.

The deficit in words is indeed huge.

SUSKIND: But frankly, quality is even more important. And Hart and Risley demonstrated differences in the quality of input, as well as many others, with children from poor language homes, hearing more prohibitions and less affirmations: “Don’t do that. Get down.” Less verbal back-and-forth, less complex vocabulary.

That combination – less vocabulary and less-complex vocabulary – creates a major detriment.

SUSKIND: By 18 months, the children from language-poor homes are processing language at six months behind those from language-rich homes. So what I always like to say — and it’s quite tragic — is that it’s not just about language, vocabulary being poured into a baby’s brain. It’s really building a fundamentally different machine.

So if a language-rich environment is really helpful in building kids’ brains, what’s to be done about it? That is what Suskind and TMW, the Thirty Million Words Initiative, are trying to figure out.

SUSKIND: We’re developing programs, evidence-based programs, that reach parents where they are to get this message out. So, we have something in the maternity wards, pediatricians’ offices. Our best-developed program is this home-visiting program where — over a six-month period of time, one-on-one with a home visitor — we go in and explain the science.

*     *     *

The main goal of the Thirty Million Words home-visiting program is to teach parents to effectively communicate with their kids during the critical years of brain development.

MICHELLE SAENZ: We’re going to be visiting Tequillia today, and her son is Sincere. He’s about 17 months or so.

That’s Michelle Saenz, the TMW home visitor. She’s outside a seven-story brick apartment building on Chicago’s West Side. The mom she’s visiting is 22 years old.

TEQUILLIA WILLIAMS: My name is Tequillia Williams and my baby’s name is Sincere Person. And we’re sitting in the living room of my apartment.

The TMW home-visit program is being run as an RCT, a randomized control trial, to test its efficacy. It’ll take a few years. Each family will get 12 visits during a six-month period, and then the researchers will do follow-up assessments for another three-and-a-half years. They’ve enrolled 200 families, and they’re hoping to have all the data by 2019. One tool the researchers use is a little device called a LENA, or Language Environment Analysis System. It’s basically a pedometer for language — it counts words.

SAENZ: OK, and so now we’re gonna look at your LENA report.

A parent is supposed to put the LENA in the kid’s t-shirt pocket so it can record how many words are spoken by both child and parent.

SAENZ: And so, on the day of your recording, Sincere heard an average of 963 words per hour. And I know for this one you had set a goal for 1,100. Really close to it, right? Just like right at there.

That’s not the only kind of data the TMW Initiative is gathering. For instance, they also measure the brain-processing speed of the kids they’re visiting.

LEFFEL: So, it’s a measure we do on two different home visits and three of our follow-up visits and it gives us insight into the children’s cognitive processing speed.

That’s Kristin Leffel, the director of operations for TMW.  She was also visiting Tequillia and Sincere.

LEFFEL: So, we set up a camera over the back of a laptop and ask the child to sit in the parent’s lap and watch a little stimulus on the computer screen. The stimulus basically puts up two images simultaneously — so it might be a fish and a puppy — and then it’ll say, “Look at the puppy. Do you see the puppy?” And then we’ll code it, frame-by-frame, to watch the child’s eye movements and it gives us insight into the processing speed in their brain.

SUSKIND: If you’re slower, that means that you’re taking a longer period of time to process a word that you already know, and you have less time to learn new  words. So, at 24 months, children who have been exposed to less language are slower processors. So they’re taking longer to understand what they already know. As a result, they can’t pick up the newer words. So it’s sort of like, you’re not just behind in the race of life, you’re a slower runner. So that’s why you don’t catch up.

DUBNER: Yeah, it’s like compound interest. I mean, it’s funny, as much as we hear the talk about how for every dollar you put away at age 20, it’s the equivalent of, you know, whatever, $50 at age 50, what you’re talking about is actually much, much more frightening to think about — the deficits at those early ages — because they start compounding at such a young age, yeah?

SUSKIND: Yeah, and it’s the reason that we need to start thinking in a preventative rather than in a remedial fashion. And it’s not to say that we have all the answers or that it’s going to be easy, but it’s about the only way that we can prevent this from occurring, because these children have potentials that we will never know. So that even when people are doing these different studies when they’re older, which are great — you never want to stop, it’s never too late — but you’ve built a different machine at that point.  

The TMW program teaches parents a simple mantra to remember how to best communicate with their kids.

SUSKIND: How you explain to families and to this country what a rich-language environment is, is complex. And what we’ve done in our research program is we’ve culled it down to what we call the three T’s: tune in, talk more, and take turns. “Tune in” is sort of the most nuanced of the T’s. It’s really seeing what your child’s interested in, following your child’s lead, using that child-directed speech, that sing-songy voice. “Talk more” is as it sounds — talking more, using rich vocabulary, narrating while you’re changing your baby’s diaper. And “take turns,” which I think is the most powerful, is really viewing your child as a conversational partner. Even before they have a word or a babble, knowing that just by responding to any glances or gestures, you’re getting them ready to have a conversation and to do it from day one. And those three T’s are the foundation and the behavioral measuring stick that we should all use in interacting with our children.

A “behavioral measuring stick” — a reminder that even though a program like Thirty Million Words is all about the words — about the cognitive skills — there’s also a lot to be gained in the non-cognitive realm, as John List and Steve Levitt saw in their Parent Academy experiment. The TMW randomized trial is now in its second round and it’s incorporating one of the key findings from the first.

SUSKIND: One of the biggest things we found is that parents, after the first trial, said, “This is great. Now I know how to build my baby’s brain and make him smart, but how do I use my language to help him behave better?” And so we actually have a week where we explain children’s executive function and prefrontal cortex development, and how their language and how their modeling can help their children’s behavior.

SAENZ: Today we’re gonna talk about another way to build Sincere’s brain and help him do well in school. And so it takes more than being smart to be successful. So, a child’s behavior is just as important as his intelligence. So, even if Sincere can count and say his ABCs, he won’t be ready to learn if he can’t sit still or follow directions.

WILLIAMS: That’s the problem I have right now. Getting him to sit still.

SAENZ: He’s pretty busy at this age, right?

WILLIAMS: Right. His green light’s on all the time.

SAENZ: And so the stronger Sincere’s red light becomes, the better he will get at self-regulating. And so a big part of helping Sincere learn to self-regulate is watching you do it. And so use your three T’s to show him how you use your red light to self-regulate in challenging situations. And so when you notice yourself feeling frustrated or upset, talk more and explain how you’re feeling to Sincere. And so tune into your tone of voice. If your tone is angry and loud, Sincere will think that it’s an appropriate way to respond.  And so if you stay calm, he will learn how to stay in control in tough situations.

DUBNER: Is there any evidence that two parents are better than one when it comes to learning language? I would think that one vigilant parent is certainly better than two crappy ones.

SUSKIND: I mean, obviously when you’re a single parent there’s a lot more stress, and there’s a lot more on your shoulders, and that’s why I always say, “When you advocate for the power of parent talk, you also have to advocate for the parent.”

DUBNER: It sounds like parent-child conversation is maybe the highest form of language acquisition. But where does TV fall into that? And is educational TV any better than the average TV?

SUSKIND: The studies are pretty compelling in the zero-to-three space that screen time is not necessarily a good thing. But in the book I talk to several studies really showing how the screen of a human being saying exactly the same thing as somebody face-to-face will have absolutely no effect on the language learning. Now with that, it’s not so much that there’s something wrong with the screen, it’s that it’s not responsively contingent. It’s not responding to the child. And I mentioned this cool study where it was a Skype with a responsively contingent adult who was responding to the child and they were able to learn vocabulary — meaning there is nothing wrong with the screen; it’s more that we’re social beings, and especially in those early, early years, you want something responding to the cues of that actual child for language to stick.

DUBNER: But presumably that could be a great use of artificial intelligence right?

SUSKIND: Absolutely.

DUBNER: I mean, if I can program whether it’s an iPad or a robot to have a conversation that is contingent on what the kid said.

SUSKIND:  Yeah, right. You know, technology right now is more on the distracting side of things. But certainly as it evolves you could imagine it becoming smart enough that maybe us parents aren’t going to be necessary in the end. But it’s got to evolve some more.

Until that evolution is a bit further along, we wanted to find out more about TV’s impact on early learning.

MELISSA KEARNEY: There’s this television program that’s been entering the homes of millions of children for decades now.

That’s Melissa Kearney.

KEARNEY: I’m a professor of economics at the University of Maryland.

She used to be a big fan of Sesame Street.

KEARNEY: My father says that my sisters and I learned everything we knew from Sesame Street.

Her favorite character?

THE COUNT: Greetings. It is I, the Count. And it is time to answer that fascinating question, what is the Sesame Street number of the day?

KEARNEY: You know, I loved the Count and I would get very excited when the Count came on. And recently it’s been pointed out to me that as an economist perhaps that’s not a surprise that I would like the guy who came on and talked about numbers.

Kearney and her colleague Phil Levine were curious about whether a show like Sesame Street really helps young kids get ready for school. Many, many researchers had already concluded that it does. There have been more than 1,000 studies on the Sesame Street effect.

But to an economist like Kearney, studies like these are potentially problematic. They often rely on self-reported data — which, as  we often preach around here, can be the lowest form of data. They may have identification and selection problems — that is, if you find that kids who watch Sesame Street do better in school than kids who don’t, how do you know that it’s not just that Sesame Street appeals to smarter kids, or to the parents of smarter kids?

KEARNEY: So, economists are really keen on finding what we call causal impacts. We don’t just want to say, “Oh, kids who watch Sesame Street when they were little did better when they got to school.” So what we need to do is figure out a way to identify kids who watch Sesame Street that had nothing to do with them or their family’s situation.

Okay, how do you do that without running a giant randomized trial where you force some kids to watch Sesame Street and forbid others to watch?

KEARNEY: Sesame Street just came on the air in 1969. So the first thing we can do is say, “Hey, relative to the kids who were just born before them, did they do better when they got to school?”

They found that yes, the generation of kids born after Sesame Street did do better in school. But that alone couldn’t prove the Sesame Street effect. Maybe schools were improving? Maybe the next generation was just smarter?

KEARNEY: And so just comparing those two generations could be misleading; it wouldn’t be due to Sesame Street.

But Kearney and Levine did find a way to make sense of the data.

KEARNEY: Good for us as researchers but bad for the kids: it turns out that a third of the kids in the country couldn’t watch Sesame Street, even when it came on, just because of limitations in television technology at the time.

In many areas, the show was broadcast on channels with a weak signal, which meant access was limited.

KEARNEY: So kids who, for example in 1969 or ’70 were living in Los Angeles or southern California more generally, or Ohio — where they lived, Sesame Street was broadcast on a weaker station and they couldn’t get that signal. So there’s this great variation for us as researchers: two-thirds of the kids in the country could watch Sesame Street, a third couldn’t. Not because of anything their parents were doing, just because of where they happened to live and the station that was broadcasting Sesame Street in their county. So using that sort of two layers of differences — the kids who were younger versus older, and the kids who lived in places where Sesame Street was on a stronger broadcast signal or weaker signal — we can identify the effect that Sesame Street had on their school readiness.

Alright then, what was the effect?

KEARNEY: We find that kids who were pre-school age in places where they could watch Sesame Street were 14 percent less likely to fall behind when they got to elementary school. If we try and make a comparison of that number to what we see in the literature studying, for example, the Head Start program, our nation’s publicly funded pre-school program, the estimated effects on school performance are very similar.

Head Start is estimated to cost roughly $8,000 per year per kid. How about Sesame Street?

KEARNEY: This costs $5 a year per kid to produce.

Now, there’s a big difference between a kid spending the day in pre-school and watching a TV show. Still, the Sesame Street effect is impressive.

KEARNEY: And in fact, that effect is entirely driven by kids who grew up in counties with higher levels of economic disadvantage. So, I mean places that had higher levels of high school drop-out, had a higher rate of single-parent households, had lower household income on average — these were the kids that really saw a relative improvement in their school performance. The effect is largest for boys and African-Americans.

[MUSIC: Brooks Williams, “Reverse Childhood”]

And this, you will remember, is opposite what John List and Steve Levitt found with their Parent Academy interventions.

LIST: Our program really, really helps Hispanic and white students and it doesn’t help blacks at all.

KEARNEY: We were surprised by the results because if you look at just a lot of studies of interventions designed to help kids, you tend to see the largest results for girls and for whites. And here we are finding the largest results are for boys and for African-Americans. So we were pretty excited about that. You know, we could speculate. This gets a little bit outside of our expertise as economists, but Sesame Street was explicitly designed to help urban kids and minority kids. So, if you remember the set from when you were a kid and watched this, it’s set on an urban stoop and African-American characters feature prominently. So it’s possible that Sesame Street did this well and urban kids and minority kids related to the characters and related to the content and maybe took more away from it than they would have if the program were designed differently.

So that’s some good news, right? Sesame Street seems to have been successful in getting kids — especially low-income, African-American boys — ready for school. But — here’s the bad news — Kearney found that, long-term, the benefit wore off.

KEARNEY: We don’t find a persistent effect on ultimate educational completion or, looking further down the road, we don’t see that large of an effect on wages or employment. So, you could think of our finding — of a strong academic elementary school outcome but not much in the long run — as being consistent with this idea that Sesame Street wasn’t impactful on those non-cognitive skills that will show up later in employment outcomes, for example.

So the evidence we’ve looked at today, from Sesame Street, from the Thirty Million Words Initiative, from the Parent Academy — as inchoate as it is — seems to suggest, at the very least, that when we talk about “learning,” we don’t really know what we’re talking about. Most of us, at least. Most of us probably think too much about cognitive skills and not enough about non-cognitive.  Most of us probably put way too much faith in the formal education system, when, in fact, the path to learning begins way before then, at home.

SUSKIND: As I’ve evolved in my thinking along this journey, it’s really about early childhood. We need to reimagine what education looks like, because we need it to follow the science. Education doesn’t start on the first day of school, it starts on the first day of life. And in this world we have made little progress on what we call the achievement gap. And I truly believe that until we address education as it should be, in a scientific and biological way — because learning begins at day one — then we’re never going to move the needle. We’re remedial rather than preventative. And that’s the larger issue.

A larger issue of course requires a larger effort. It requires that rather than simply worrying about the success of your kids, or my kids, we might do better to think about a new, collective effort. To change our expectations of what it means to be a society that truly values learning. This would require a lot of work, perhaps even a lot of sacrifice.

Dana Suskind has three kids of her own: 18, 16, and 13 years old. Her book, Thirty Million Words, ends with a terrible story, something that happened several years ago at Lake Michigan. “Our three children,” she writes, “were playing in the sand, watched over by my husband, their father, Don Liu. As he stood at the shoreline he suddenly noticed in the turbulent, chaotic distance two young boys struggling in the raging waters. He started running into the lake as our younger daughter cried out, ‘Dad, don’t go!’ They were the last words she ever said to her father. The two boys got back alive. My husband, always fearless when it came to helping others, died, overwhelmed by the torrential pounding of the waves and the gripping undertow.”

DUBNER: Now, your husband was a pediatrician as well, correct? A surgeon?

SUSKIND: He was a pediatric surgeon. He did, actually, mostly minimally invasive surgery. So, with little portholes. He was a spectacular surgeon. And he said it was because he was really great at video games.

DUBNER: You’re making an argument against yourself now, you understand.

SUSKIND: I know. Well, it was terrible. We would be next to each other at night, and he’d be playing video games and I’d be writing a paper. But then he always made everything look easy. So he’d be the one who got the grants. And so yeah, he was just awesome. And I don’t know if you actually got to the end of the book.

DUBNER: I did.

SUSKIND: But he’s the epilogue. And for me, I included him both so he could continue on this journey as well as to, really — he’s sort of a metaphor for what we need to be thinking about in this country. That every child in this country is our own and we should want for them what we want for our own children.

*     *     *

Freakonomics Radio is produced by WNYC Studios and Dubner Productions. This episode was produced by Greg Rosalsky. Our staff also includes Alison Hockenberry, Merritt Jacob, Max Miller, Stephanie Tam, Vera Carothers, Harry Huggins and Brian Gutierrez. You can subscribe to Freakonomics Radio on Apple Podcasts, or wherever you get your podcasts. You can also find us on Twitter, Facebook, or via email at radio@freakonomics.com.