Search the Site

Episode Transcript

Have you ever had a question about how Freakonomics Radio gets made? If so, we want to hear from you — any question at all, about how we choose our topics, why we interview who we do, how the whole process works. Stephen J. Dubner will be sitting down to answer your questions with Alison Craiglow, our executive producer, for a series of episodes called “The Hidden Side of Freakonomics Radio,” which we’ll put out on Stitcher Premium. That’s the subscription service where you can listen to our entire archive ad-free and hear bonus material like this. To submit a question, send an e-mail to, with the subject line “Hidden Side.”

*      *      *

If you’ve ever had a child, or ever been a child, you know there’s a lot of parenting advice out there. Much of which is not very nuanced.

Emily OSTER: No one’s in the middle. People are yelling. The first person is like, “Well, I did that, and my kid’s amazing.” And then someone will be like, “Well, actually, if you do that, there’s a very good chance your baby will die, and only someone who hates their baby would do that.”

Some parenting decisions are controversial, and we hear about them all the time:

OSTER: Vaccines do not cause autism. We have a tremendous amount of data showing that that is not true.

But there are many other decisions that don’t get much scrutiny.

OSTER: And the guidelines seemed really arbitrary.

Like the restrictions on certain foods during pregnancy — and alcohol and caffeine. Whether or not to let the baby “cry it out.” And the long-term effects of: day care; of screen time; of eating nothing but buttered noodles for the first 10 years of life. Wouldn’t it be nice if someone out there could cut through the dogma and the old wives’ tales and use data to help parents make decisions?

OSTER: And so I really started digging into, “Well actually, what should we do here?”

*      *      *

We’re speaking today with Emily Oster.

OSTER: I am a professor of economics and public affairs at Brown University.

DUBNER: You’re also, we should say, married to another Brown economist, Jesse Shapiro, yes?

OSTER: Yes, I am. That’s my husband.

DUBNER: We should also say your parents are also both economists.

OSTER: Yes, that’s also true.

DUBNER: Can you talk just for a second about the degree of inbreeding among economists, and whether it’s typical in the social sciences or academia writ large?

OSTER: So it’s interesting. I think a lot of female economists are married to male economists. There aren’t as many female economists, so some of the male economists are not able to marry female economists. And whether that’s as common in other academic fields, I’m not sure. But if you like it, it’s great. I think it’s pretty fun to get to do work stuff in addition to home stuff with my spouse.

A lot of Oster’s research has been related to healthcare.

OSTER: Studying people’s health behaviors and trying to understand why people don’t always behave in the ways that our economic models would suggest. So, why don’t people always seek out information about their health? Studying people who are at risk for Huntington’s Disease and why they don’t always get genetic testing. Looking at why people don’t seem to always diet when we think that they should. Or what kind of health recommendations people do or do not respond to.

What sort of tools does Oster use for this kind of research?

OSTER: There is a lot of emphasis on big data sets, using administrative data, looking at questions that can be well-answered in those data.

So several years ago, when Oster was pregnant with her first child, she naturally went looking for good data to help her make good decisions.

OSTER: And even around something like prenatal testing, trying to understand, should I have this screening test for genetic disorders, or this other screening test? The guidelines seemed really arbitrary, and to be based on age and not on anything about preferences — which is totally outside of how I think about decision-making.

And so I really started digging into like, okay, well actually, what should we do here? And what I came up with suggested some of the data was really flawed that people were using for recommendations. The recommendations didn’t even really make any sense, even given the flawed data that they were using. It isn’t like there’s some secret repository of knowledge about babies that you’re missing out on.

Oster went back and started reading the underlying studies that contribute to the conventional wisdom on pregnancy and child-rearing. She found that a lot of the studies were built around small sample sizes or incomplete data. A lot of the analysis didn’t control for things like income and education level. Consider, for instance, one of the most controversial topics of early motherhood: breastfeeding.

OSTER: So higher-income women, particularly in the U.S., women with more education, they’re more likely to breastfeed. And so if later you look, and their kids are doing better in school or are thinner — those are things which are also correlated with parental education or parental income, with resources, that the family has. And so it’s really hard to learn from that comparison about the real causal relationships. And that comes up all the time — not just in breastfeeding, in everything.

So Oster read the breastfeeding studies, including one that was based on a large randomized trial. What’d she conclude?

OSTER: There are some small but not zero benefits in the short run, particularly around improving digestive health, lowering episodes of diarrhea, and maybe some evidence that it lowers rates of ear infections in the first year of life. But many of these claims that people make — breastfeeding is going to give your kid an I.Q. bump, breastfeeding is going to make your kid thin, it’s going to prevent allergies or asthma later — these things are just not supported in the data.

To say something is “not supported in the data” doesn’t mean it’s not true. Just that when people say it is true, their argument is more likely to be based on some sort of wish, or belief, as opposed to scientific evidence. Oster did find some compelling evidence about breastfeeding, but not as a benefit to the baby.

OSTER: Yes, so this is the one long-term effect where it looks like maybe we have some good evidence, it suggests that it may actually lower breast cancer risks for the mother.

DUBNER: What’s the mechanism for that?

OSTER: Mechanisms are always hard, but in this case, I think we have some sense that it changes some of the composition of the cells in the breast in a way that may help protect against breast cancer.

DUBNER: Okay. How real is nipple confusion? The idea that if you feed your baby with a bottle, it will get used to that nipple and then be confused if you try to breastfeed later?

OSTER: Nipple confusion is made up. So, particularly around something like pacifiers — you don’t give your kid a pacifier because they won’t nurse — there is just no evidence for that.

DUBNER: So even a not-very-brain-developed baby—

OSTER: —is able to differentiate between a breast and a pacifier. That’s right. Amazing.

DUBNER: Okay, what about dietary restrictions for a breastfeeding mother?

OSTER: Mostly none. And the question people ask me all the time is like, “Is it okay to drink while I am breastfeeding?” And the answer is, not like a sailor, but some, yes. Totally fine. The concentrations are very low in breast milk. Caffeine, fine. Some babies are very sensitive to some things like caffeine, so if you find that you drink a cup of coffee and you nurse your kid and they get totally crazy, then you may need to adjust.

DUBNER: What about medication, especially antidepressants?

OSTER: Antidepressants do pass through breast milk to the baby. And so this is something that women have to talk to their doctors about. But in general, many antidepressants are safe for use, and also postpartum depression is a very significant issue. And if you need treatment, you need treatment. And that is something that should be paramount.

DUBNER: So, on balance, the benefits of breastfeeding are what?

OSTER: I think you want to imagine that there are some small benefits to breastfeeding in the short term, and those may be enough to try. And also I should say a lot of women enjoy breastfeeding, and they find it to be a nice way to bond with their infant, and that is of course a great reason to do it.

I think the thing that is too bad and is not great is when people sort of build up breastfeeding in their mind as like, “This is the only way to give my baby the best start. And if I don’t do this, that’s giving them a bad start.” And that just isn’t true.

As Oster sees it, one big problem with parenting is that many of the conversations instantly devolve into a level of partisanship that can make our political discourse seem courtly.

OSTER: For me, the thing that really encapsulates this is the Facebook conversations. Somebody will ask a question like, “My three-week-old baby is not sleeping well, and I’m thinking about keeping them in the bed with me, what do you ladies think about this?”

Pretty soon, the shouting starts.

OSTER And then somebody will be like, “Okay, we have to shut down this post because this is too much.” And some of what I try to do is push against that, and say, you could each make different choices and they could both be right. And just because it’s not the same choice doesn’t mean it’s wrong.

Economists believe in preferences — and they also believe it’s perfectly sensible for different people to have different preferences, as long as they’re making decisions with a full knowledge of the costs and the benefits. And the knowledge that most decisions do have both costs and benefits. And yet somehow, when it comes to parenting:

OSTER: I think that there is a knee-jerk to be like, “Well, if anyone ever said that this might be dangerous, no one should ever do it, ever.” I think that there is sometimes a discomfort with facing up to evidence and also to the uncertainties that come with data, that lead doctors, medical professionals, medical organizations, to want to make more blanket statements than are always appropriate, and to be less comfortable with explaining nuance to their patients than they might otherwise be.

Oster had no such discomfort with nuance. She set out to explore the parenting terrain using data as her guide. The result has been two books. The first, published a few years ago, is called Expecting Better: Why the Conventional Pregnancy Wisdom Is Wrong — and What You Really Need to Know. The new book is called Cribsheet: A Data-Driven Guide to Better, More Relaxed Parenting, from Birth to PreschoolOster appreciates that there are systemic reasons for the medical field to be cautious: remember: first, do no harm; also, there’s the threat of a malpractice suit. But Oster wanted to think about risk rationally — not as a doctor, hoping to avoid liability; or even as a parent, wanting nothing bad to ever happen to her children. Instead, she just wanted to think about risk as an economist.

OSTER: First of all, let’s interrogate a little bit whether those risks are really real, and are really significant. And then also to interrogate you have to trade off the risks maybe against some other benefits. And in something like pregnancy, you think about treating really severe nausea. There’s this “Oh, don’t take anything for that, just suffer through it.” So actually, that can be really debilitating. And it may make sense for people to take something even if we are not 1,000 percent sure that there are absolutely no risks to it, because it may outweigh some other risks. And I think we sometimes forget that.

DUBNER: And what about facing head-on risks that you’re describing as relatively small while totally ignoring other, let’s say, daily risks that are actually relatively large, like getting in a car?

OSTER: I am constantly comparing things to getting in a car, because getting in a car is very risky. And I think that there are many kinds of risks that people talk about in pregnancy and childhood which are far less risky than getting in a car, where people are like, “Oh, only somebody who’s a terrible parent would even consider doing that.” It’s like, “Well, actually, do you get in the car?”

When it comes to advising parents on risks, one problem Oster identifies is that the advisors — doctors, primarily — aren’t necessarily practiced in risk-reward calculations.

OSTER: There is relatively little training on data analysis in medical school. And now, that of course does not mean that doctors are not data-literate. Many of the doctors that I know are very data-literate and think very carefully about these kind of issues. But it is true that generally this is not a training that gets much play in medical school.

I think there are some good reasons for that — which is, there are many things about being a doctor that are about doing things correctly and understanding how the biology works, which are much more important. And you can’t teach everybody everything.

“You can’t teach everybody everything.” Fair enough. But if there’s one group of people in the world who think they can learn just about anything, it’s economists. This is a long-standing complaint among other academics and scientists and assorted smart people. The field of economics does carry an air of triumphalism; many economists feel they can contribute insights to areas that lay well outside their own expertise — areas like education, criminology, medicine.

OSTER: There’s also a lot about decision theory, and about how do you structure a decision in a way that helps you make a good choice? And that’s really economics. Developmental psychology and obstetrics and pediatrics are not sciences of decision-making. And so I think especially around the issues where you’ve got to think about what’s the best for your family, you need someone doing decision science.

There’s something to be said for this: economists have analytical tools that are useful on many topics; they’re really good at working with very large data sets; and there can be a big upside in having an outsider’s perspective on hard problems. But economists’ triumphalism — or maybe you’d call it colonialism — it also has its downsides. As Emily Oster discovered first-hand. Years ago, she was trying to understand why the ratio of males to females was so off-kilter in many places, especially Asia. Most previous explanations pointed to violence against women and girls; or the selective abortion of female fetuses; or even infanticide. Oster offered another explanation.

OSTER: I wrote a paper in graduate school which argued that parents who are carriers of the hepatitis B virus have more male children, and then this explains some gender imbalances.

The theory was that a pregnant woman with hepatitis B was much more likely to give birth to a boy than a girl — although the mechanism wasn’t clear; it could have been that female fetuses were more likely to be miscarried when exposed to the hepatitis B virus.

OSTER: And then subsequently some research came out which suggested that basic fact was not true in the data. And then I did some subsequent follow-up research, which also showed that that was not true in the data.

She had to walk back her earlier conclusion. Which had gotten a lot of attention. Including from the authors of Freakonomics. We too walked back her conclusion.

OSTER: That mistake, that error, that episode, has had a big impact on how I think about my work and how I think about the importance of being careful. And so I try to be careful.

DUBNER: Did it kind of make you feel that the whole goal of establishing causality was much harder than you used to think it was?

OSTER: Yeah, I think it taught me a little bit to be more cautious about some of the— I thought I had a really good set of causal evidence around this problem, and then it turned out not to be right. And of course, sometimes things are not right. And so I think that it did give me pause about some aspects of causality.

It is probably a good idea to take some pause about causality. Causality’s often much harder to establish than it might first appear. Especially when the data aren’t abundant. Especially when the topic is something as universal — and controversial — as parenting.

OSTER: Some kinds of things people tell you are just completely made-up, old wives’ tales — like if your belly sticks out to the front, that’s a boy. And then there are some things where it’ll be food restrictions, where the answer is the restrictions come from data, but there’s a wide variety of quality in how good the data is and how good the conclusions are.

Okay, so: let’s talk about some of these conclusions — the conventional wisdom — and how solid it is or isn’t. Let’s start with pregnancy. And perhaps the most obvious don’t from the do-and-don’t list:

OSTER: In the U.S. there is a blanket no-alcohol during pregnancy, even a small amount of alcohol can be dangerous.

And having read the underlying studies, what did Oster conclude?

OSTER: It is definitely true that drinking a lot of alcohol is very bad, and even one or two times having a large amount of alcohol can be very dangerous. But the data does not support the conclusion that occasional alcohol consumption — say, no more than a glass at a time, a few times a week — is dangerous for your baby.

DUBNER: And you made a lot of new friends by writing that, did you not?

OSTER: So many friends, yes. Not everyone was very happy with that.

The National Organization on Fetal Alcohol Syndrome called Oster’s conclusions “deeply flawed and harmful.” But Oster stands by her conclusion, and it seems as if the obstetrics community is moving in her direction.

OSTER: The truth is, about half of the obstetricians in the U.S. say that they tell their patients it’s fine to have an occasional glass of wine. And so my guess is that more people listened to that after reading my book than before.

DUBNER: What about caffeine?

OSTER: So, caffeine, again, the restrictions are very, very stringent. And I think some people take that to mean none, no caffeine. The truth is, there’s certainly no evidence that having two cups of coffee a day is dangerous. And there really isn’t much evidence that going up to, say, three or four cups a day, has any negative impact either. When you get into eight cups a day, that data is a little more complicated.

DUBNER: There is a long list of foods that some pregnant women avoid.

OSTER: Yes. No deli meats, no soft cheeses, no sushi, etc. There are a few things that you should avoid. Probably deli turkey, things that sit around in a steam table, not so good. Probably unpasteurized soft cheese, also worth avoiding. But many of these things — sushi, ham — that are on the restricted list, most women are likely to look at the evidence and think it’s actually fine.

DUBNER: And why deli turkey but not other deli meats?

OSTER: Deli turkey is more likely to harbor listeria than other deli meats.

DUBNER: And then talk about smoking and nicotine. I want you to handle them separately, if known. Smoking, it sounds as though it’s pretty indisputably bad.

OSTER: Smoking is bad. Smoking is particularly bad for birthweight, and the evidence for that is pretty good.

DUBNER: And we should say, low birthweight is a very good proxy for baby health generally, yes?

OSTER: Yes. It is generally the good proxy we use for that.

DUBNER: But then, what do you know about nicotine? Because obviously there are other ways to deliver nicotine, and there are those who argue that nicotine itself is actually a pretty nifty drug, in moderation.

OSTER: Yeah, I think the issue is, we don’t actually know that much about how should we think about nicotine-replacement therapy as relative to cigarettes — or e-cigarettes relative to cigarettes. Is just hasn’t been studied much.

Note to self: that would be a good topic for another episode — the risks of nicotine itself — now that vaping has become so popular. Anyway: back to the risks, and alleged risks, surrounding pregnancy and childbirth. One of the most controversial topics around childbirth is C-section versus vaginal delivery.

OSTER: Yes. And I think what we know there is that basically in the short run, the recovery from a vaginal delivery is on average a bit easier than a C-section, so women tend to be up and about a few weeks faster. In the long run, actually, recovery is very similar.

It is also true, though, that for later deliveries, having had an earlier C-section can increase some complications. So, people sometimes ask, “Well, is there really any downside to having a C-section?” I think the answer depends a lot on whether you want to have more kids. If you want to have more kids, the downsides are more salient than if you’re done.

DUBNER: And what about outcome on the babies, whether cognitive, physiological, whatever?

OSTER: We just don’t have any evidence suggesting there’s any differences in outcomes at all. People say things like, “It’s important to have the microbiome, and you have to rub the vaginal secretions on the baby.” We just don’t have any evidence that that works or not.

DUBNER: And when you say we don’t have any evidence, meaning there’s not enough evidence really to think clearly about it, or there is quite a bit of evidence and it just doesn’t suggest a difference?

OSTER: I think that’s an important distinction — I would say this is more in the category of, we just don’t have enough evidence. And so the evidence that we do have doesn’t suggest large differences. And in the general question, about many of the kinds of outcomes, like survival and so on, we have good evidence that it doesn’t matter. On some of these more subtle things like disease resistance, allergy, immunity, I don’t think we have great evidence.

You can start to see why so many parents, or would-be parents, get so confused by the avalanche of information coming at them. Information that’s often not very well-sourced, or that’s got an agenda attached to it; or — and this is pretty common — information that used to argue for one decision and now argues for the exact opposite. There’s a 2003 book, by Anne Hulbert, called Raising America: Experts, Parents, and a Century of Advice About Children. It does a great job showing how many flip-flops there’ve been over time.

Consider, for instance, how parents today are urged to engage and intellectually stimulate their children. And there’s a lot of evidence that things like the early acquisition of language is incredibly powerful. But in the early 20th century, one of the most renowned pediatricians of his time, L. Emmett Holt, cautioned that a baby is not a “plaything” and there should be “no forcing, no pressure, no undue stimulation” during the first two years of life. Holt’s argument was that the brain was growing so vigorously during that period that overstimulation might cause “a great deal of harm.” He also believed a baby should be left to cry for 15 to 30 minutes a day. “It is the baby’s exercise,” he wrote.

*      *      *

Some people argue the world is in such a state that it’s irresponsible to bring any more humans into it. There’s a British movement called BirthStrike, for instance, composed primarily of women who have declared their intention to “not bear children due to the severity of the ecological crisis and the current inaction of governing forces in the face of this existential threat.”

Indeed, the global fertility rate is about half of what it was in 1962, which turns out to have been the peak year for global population growth. The fertility rate is particularly low in wealthy places like Japan and many western European countries; the U.S. rate is somewhat higher. But overall, the global population continues to grow: roughly seven-and-a-half billion people today, up from less than four-and-a-half billion in 1980, 3 billion in 1960, and well under 2 billion in 1900. A few years ago, for an episode called “Why Do People Keep Having Children,” I asked Emily Oster that question.

OSTER: I think this is probably an open question for debate. I think many people would tell you that it’s the biological imperative. I think that some people would tell you, kids are enjoyable. I think some people would tell you, particularly in developing countries, people have kids as an investment in their old age or even to work on their farms when the kids are young. So, I think those are probably the leading-candidate explanations.

That, at least, is how an economist like Oster sees it. She and her husband have two kids: a daughter, Penelope, who’s 8; and a son named Finn, who’s 4.

OSTER: And they are great.

But the scarcity of data about parenting led Oster to seek out what she could find, and write it up in two books: Expecting Better, about pregnancy; and now Cribsheet, which essentially says: okay, you’ve had your baby — now what?

OSTER: I think one of the biggest challenges that we face when we parent now is the perception that if you are doing something for yourself, it must necessarily not help your baby or your small child. And I think so much of the rhetoric around this almost martyr-like approach to some aspects of parenting, it’s like, “Well, I haven’t slept in three years. But that’s because I love my baby.” And I think it should be fine to say, “Look, I don’t care if my kid doesn’t sleep because I like getting up in the middle of the night with them.” Some people will say that, I think that’s totally fine. But the idea that somehow that is the thing that makes you a good parent is something that I think we should move away from. And really, rather than thinking about somehow, all of my sacrifices are how I prove I love my kid, you could just love your kid and also sleep. That’s also fine.

DUBNER: So, babies and little kids sleep a lot. And I’m guessing that parents have many questions and confusions about the do’s and don’ts of baby sleeping. Why don’t you tell us what are some of the big issues that people have, and then how they should be thinking about it.

OSTER: The two biggest questions in sleep are where the baby should sleep and whether you should let them, what is colloquially called, “cry it out,” whether you should do some sleep training with them. So, on the first question, the issue is should you let your baby sleep in your bed? And a lot of people are tempted to do that because actually many babies sleep better in the bed. And then if you need to breastfeed in the middle of the night, or do something else, it is easier to roll over and just not have to get out of bed and get on your bathrobe and go down the hall and get your baby. So that’s the plus. But there are pretty strong restrictions on, you know, you shouldn’t have your baby sleeping in the bed with you, because you could roll over them and that’s a risk factor for S.I.D.S.

S.I.D.S. stands for “sudden-infant death syndrome.”

OSTER: This is a sort of general term for the condition in which an infant dies in the crib or in the bed, without other obvious risk factors.

The American Academy of Pediatrics says babies up to one year old should sleep on their backs on a firm surface, without any pillows or bedding.

OSTER: Yes, sleeping on the back is a good idea. The Back-to-Sleep Campaign has been very good at preventing S.I.D.S.

DUBNER: Okay, now, that said, for people of a certain age, they will remember that the advice on this has flip-flopped a lot of times. How persuaded are you that sleeping on the back is the definitive good idea?

OSTER: I was pretty persuaded. I mean, I came into this sort of thinking, I wonder if this is one of these many things that is not super-supported. But actually because of these flip-flops, we actually have some pretty good evidence. So for example, you can see in places that have flip-flopped, the death rates also flip-flop, suggesting that back-sleeping is very important.

The American Academy of Pediatrics also suggests that babies sleep in the same room as the parents but not sleep in the parents’ bed.

OSTER: The degree to which that is a risk depends a lot on the other kinds of behaviors that you’re engaged in.

DUBNER: Meaning smoking and drinking primarily?

OSTER: Smoking and drinking, yes. If smoke is around that baby, that’s already compromising their respiration and because that kind of compromise is also a risk for S.I.D.S., that’s the issue there. And for drinking, it’s simpler — if you’re intoxicated, you are more likely to roll over on the baby and not notice.

DUBNER: So, basically a baby in the bed, if there’s not smoking and drinking going on, is very low-risk, correct?

OSTER: It’s low-risk. Most of the evidence would suggest that there is some risk to that, but it’s small.

DUBNER: Apparently there is a fair amount of sleep-sharing, meaning parent and baby sleeping together, on a sofa versus a bed. How bad is that, if so?

OSTER: That is extremely dangerous. Of all of these things, sleeping on a sofa with your infant is something you should not do.

DUBNER: Why? I mean, how do bad things happen there?

OSTER: People fall asleep sitting up on the sofa with their baby because they are trying to stay awake and not fall asleep in bed with their baby, or some other reason. And then the baby falls over, the sofa is very soft. People fall over, they fall over on the baby, the risk there is suffocation. Yes, that’s really risky.

DUBNER: So the way you just told that story, it’s the desired avoidance of co-sleeping in a bed that leads to the bad thing, yes?

OSTER: Yes. And I think that’s something that doesn’t get enough— people are really trying very hard to stay awake, but of course you’re exhausted, and I think we’d be better off telling people that they would be better off sleeping safely in a bed than sleeping on a sofa, for sure.

DUBNER: So you’re a super-smart person, and I’m curious to know how you would describe the caliber of your thinking during that haze of, let’s say, the first year of your first kid.

OSTER: I mean, just terrible. I think part of it was I was constantly trying to figure out what was the right thing to do. Not so much around these data things, but just what exactly works for your kid. There’s this tendency to be looking for patterns all the time. Like, “Okay, they slept for six hours. What did I do? Was it like this particular song? I think you should probably sing that song again.”

DUBNER: And then what about sleep training and “crying it out” or other methods?

OSTER: Here the thing that people will tell you is that if you do this, your kid will be forever damaged and unable to form adult relationships. There isn’t any evidence for that. There actually is a lot of randomized trial evidence on the impacts of sleep-training programs on infant sleep, and there’s just no evidence to suggest that sleep training has any negative consequences. It does make your kid sleep better.

DUBNER: So if letting a baby cry it out leads to better sleep and baby sleeping presumably leads to better maternal sleep and maybe paternal sleep, I’m curious about the relationship between the baby’s sleep and maternal depression, and whether the cost of letting your kid cry it out might be really, really, really worth it in the long term, to the parents especially.

OSTER: Yes. What’s interesting about these studies of baby sleep is the main outcome they’re interested in is maternal depression, or parental functioning. And you actually see in the randomized data that one of the outcomes of doing a sleep-training program with your kid is lowered maternal depression. And some of those effects are really big, because sleep deprivation, we know it’s very hard to look positively at your life when you’re exhausted. And people whose kids really don’t sleep well, depression is a very significant risk.

And so I think that that gets lost a little bit in some of these discussions, is that there are some real benefits. It’s not just something you selfishly do because you’re hoping to go out to the club.

DUBNER: Let’s talk about the big landmark events for children that parents watch out for: walking, talking, and so on. What would you say is the single biggest misperception about those landmark events?

OSTER: Those physical milestones, there is a very wide range of normal. I think much wider than people perceive. And being on one or the other end of the ranges of normal is not worse. So, kids who walk late are no less likely to, say, be able to walk later or have lower I.Q., or anything like that.

People get very focused on physical milestones. And if they’re walking early, “maybe they’re going to go to the Olympics!” Your kid’s not going to the Olympics, probably.

DUBNER: What can you tell us about kids and germs and the hygiene hypothesis?

OSTER: So the hygiene hypothesis refers to the idea that you should expose your kids to germs, because then they will be healthier later and they will have fewer allergic reactions. I think that there’s some evidence to suggest that that’s true. And so for that reason, when your kid gets sick when they’re toddler age, usually we don’t worry too much. And you probably don’t need to be super-obsessive about never exposing them to any germs.

When your baby is very, very little, it actually is a good idea to avoid germs because if they get sick, then it sort of sets off a cascade of interventions which will happen even if they just have a cold, which you want to avoid.

DUBNER: But is exposure to germs, let’s say, in toddlerhood and up, is it actually long-term beneficial, then?

OSTER: I think to the extent we have evidence, it suggests it probably is somewhat long-term beneficial, at least in preventing them from getting sick later.

DUBNER: Okay, same questions, then, about allergens, because there’s certainly been a lot more attention paid to allergies — peanuts is maybe the most famous one. It would seem to be there’s a big spike in these. But it may just be these were previously undiagnosed. What can you tell us about exposure to allergens and long-term costs or benefits?

OSTER: Yes, so this is one of the biggest changes even in the last five or ten years, has been the recognition that the best way to prevent your kid from developing allergies to things like peanuts and eggs and wheat is to give it to them when they are little, not to avoid it.

And so people were told, “Don’t give your kids peanuts until they’re two, because it could be an allergen.” It turns out that is a great way to produce allergies in people, and a good way to prevent allergies is to give them peanuts very early on, so that advice has totally switched.

DUBNER: I can imagine thousands of parents listening to you right now and shuddering with the idea that, “Oh, the last thing I’m going to do is give my kids peanut butter when they’re three months old, because I barely got used to the kid, and now I’m endangering him or her.” So how do you do that early exposure while protecting yourself against possible downside?

OSTER: If you’re very anxious about this, sometimes they’ll tell people, bring your kid to the E.R. and give them some peanut snacks. You know, unless your family has a high risk of allergies, unless you have a reason to think your kid is at high risk of allergies, they almost certainly are not going to have that kind of reaction. It’s probably not something to actually worry that much about.

DUBNER: So what do you know about the incidence of, let’s say, peanut allergies now versus 50 years ago? Is it truly higher?

OSTER: It has gone up.

DUBNER: And why is that? Because I would think that more kids would have access to peanuts earlier now than they used to.

OSTER: There was a long period in which they told people not to expose people to peanuts. So I think that was not great.

DUBNER: So the supposed prevention was part of the problem, you’re saying.

OSTER: I believe that it’s part of the problem, yes.

DUBNER: Gotcha. Okay, let’s talk about vaccinations, which didn’t used to be very controversial but has become so in the last maybe 10 or 15 years. Talk about, I guess, the controversy, the beliefs, and where you feel the evidence lies.

OSTER: Childhood vaccinations are designed to prevent diseases, like pertussis or measles. There has been a lot of discussion in the last two decades about the possibility that vaccines cause autism or other kinds of negative consequences. There was a very, very damaging paper by a guy named Andrew Wakefield in The Lancetwhich suggested that the measles, mumps, and rubella vaccine contributed to autism. It turns out that was not only wrong, but also completely made-up, fraudulent. He lost his medical license. But still, those concerns have been really, really prevalent, and have contributed to lower rates of vaccination.

There is no evidence for those negative consequences of vaccines. Vaccines do not cause autism. We have a tremendous amount of data showing that that is not true. They do not cause autism. The kinds of things people cite as risks are simply not there. And also vaccines do prevent disease.

DUBNER: What are the greatest downsides of the suite of vaccines that are commonly applied?

OSTER: The biggest risk of vaccines — and I should say, I am trying hard to actually take seriously the concerns that people have about these. I mean, I think it’s very clear you should vaccinate your kid, but I think we also do people a disservice by not explaining to them what is the real truth about the risks, which turn out to be very small.

So the big thing is that when you vaccinate your kid, they may get a fever. And that’s a very common reaction to the measles, mumps, and rubella vaccine in particular, that it’s an immune challenge. A very small share of kids, if you get a fever very fast, can have a seizure, which is very scary, but actually also has no long-term consequences and is, again, a very rare complication. And there is a small number of other things that can happen, say, to a severely immune-compromised child which are cited as risks of vaccines. But of course, if your kid is very immunocompromised they will not be vaccinated.

DUBNER: Talk about discipline for a minute, and I’m especially curious about the value of consistency.

OSTER: When we look at evidence on discipline, the consistency emerges as the most important thing. So there’s a lot of different strategies. But in almost all cases, it’s just, you want to pick a strategy and stick with it, because you want your kid to know what to expect. Like, if you say, “If you don’t put down that toy, the following thing is going to happen,” you have to make sure that thing happens.

Which is why, if you’re on an airplane with your kid, and you say, “If you don’t stop kicking the chair, I’m going to leave you on the airplane,” that’s actually not a good threat because you’re not going to leave them on the airplane.

DUBNER: Says you.

OSTER: Yeah, unless you’re prepared to leave them on the airplane, don’t say that.

DUBNER: And then, how great is spanking? I assume it’s wildly effective?

OSTER: Spanking is not effective. There is no evidence that it is good, and a fair amount of evidence that it’s bad.

DUBNER: And how does that bad evidence manifest itself?

OSTER: That when you spank a kid, there are more behavior problems later.

DUBNER: Can you talk for a second about the tradeoff between a parent wanting their kid to be happy, and safe also, and wanting to do what’s long-term best for them, when it comes to child-rearing?

OSTER: This is a very hard thing. And I think it comes up even in something like discipline, where in the moment, you don’t want your kid to be sad. You don’t want to punish them. But you need to do that because that’s how they learn how to behave correctly and how to be a successful adult. But it’s hard, because you love your kids, you want them to be happy all the time.

DUBNER: Do you have any advice for people who are really torn between those two poles?


DUBNER: That’s quite hopeless.

OSTER: No, that’s not true. So, when you do some of these sleep-training things with your kid, it can be very hard. People don’t like to listen to their kids cry. And I have a WhatsApp channel with my best friends and I’ll be like, “Okay, I’m doing this. Can you please tell me that, like, it’s okay?” And they’ll be like, “Yes, you can do it! It’s great.”

I harp a lot on like, moms can be mean to each other and parents can be judgmental. But of course there’s also a camaraderie that comes with parenting that’s really special and can be really, really important for surviving.

DUBNER: Talk a little bit about media consumption — TV and/or screens and all the media that those can deliver. What’s the sensible way for modern parents to think about, again, the upsides and downsides?

OSTER: So, this is another place where evidence is not very good. On the one hand, it is very hard to imagine that having your kid watch a half an hour of TV a week while you take a shower is going to be bad for them. And indeed, the evidence would suggest that that is fine. It is also the case that if your kid is only watching TV all the time, we will generally think that is probably not good.

Some of the questions you would be more interested in answering are, what about 90 minutes a day, or a couple of hours a day, is that a problem? We do have some evidence on that — some of it’s actually by my husband, oddly — which suggests that watching TV in that range of time is not damaging, doesn’t have impacts on later I.Q. or test scores.

What we’re missing, I think, is answers to questions like, what about iPad games, what about screen time on the phone? And that just hasn’t been very well-studied because those technologies are relatively new.

DUBNER: So, in terms of figuring out how the baby or kid will be cared for, and really who’s going to take care of the child, you advocate creating a decision tree. What share of your readers do you think will actually create a decision tree and follow it?

OSTER: One hundred percent, obviously. No, a lot of what I’m trying to do is just say, look, there’s a framework, a way that you can think about these decisions, and just structuring the way you think about it a little bit can sometimes be very helpful to come to the choice that works best for you.

DUBNER: What do you know about the relationship between daycare and cognitive and behavioral outcomes?

OSTER: Again, it’s very hard to study, and to the extent that we see data, it suggests that maybe there’s a little bit of negative impacts on some behavior from very early on and maybe some positive impacts from having your kid in daycare later. So if you summed it up, it’s like, if you said, “I’m going to either have a nanny ‘til school, or I’m going to have daycare,” it’s basically a wash. So do what works for you.

DUBNER: Are kids who go to daycare less attached to their parents?

OSTER: No. They are not.

DUBNER: You write the following about going back to work yourself, as an economist: “The eighth hour at my job is better than the fifth hour with the kids on a typical day. And that is why I have a job, because I like it. It should be okay to say this.” Is it not okay to say this among certain friends or family members?

OSTER: Yeah, I think sometimes it’s not. I think it can feel often in these conversations that parents have with each other, and particularly moms, that if you work, you’re supposed to say, “Oh, well of course I have to work. I’d love to spend more time with my kids.” And if you don’t work, you sort of say, “Well, I have to stay home.” And I think you get a lot of judgment on both sides in a way that I think is really not helpful.

DUBNER: Well, what can you tell us about the benefits of a stay-at-home parent?

OSTER: There really aren’t many differences across kids. Depending on how the sort of work configuration of the household. I should say, this is one of those examples where it’s very hard to study, because whether parents choose to work is not assigned randomly, it correlates with income. It correlates with all that stuff.

But to the extent that we know, we certainly don’t have any good evidence suggesting that one kind of parenting, whether it’s stay-at-home or or not stay home, that one of those is better than the other.

DUBNER: You write that the U.S. has subpar maternity-leave policies, and you compare that unfavorably to European and other countries. So what effect do those relatively long-term parental-leave policies have on kids’ outcomes?

OSTER: Yes, so I want to distinguish between two things. One is going from, say, no maternity-leave policy, which is what a lot of people have in the U.S., to something like four months, which would be the low end of what you would get in these European countries. And I think there, we do have good evidence from policy changes in the U.S. and from other places that that is good for infant health, that having some maternity-leave opportunities, some ability to be home early in life, is good.

When we then think about going from four months to, say, two years, there we don’t see much evidence that that influences long-term outcomes. I think that going to everybody having a year is probably not as important as trying to make sure that people have a few months.

DUBNER: You write the following: “By the time I had Penelope” — which was your first kid — “I was 31. Up to that point in my life, there had been surprisingly few instances in which I could not defeat a problem with hard work.” So, how did the problems of parenting differ? Were your work ethic and intellect as effective in this realm? And I’m curious how it played out.

OSTER: I think the introduction of another person really limits the way that work can help. And so I think I wrote that in the context of thinking about infant crying, and these moments with your baby early on, where you think, “Okay, why don’t you just get on the breast and nurse?” They just won’t do it. And it doesn’t matter how hard you work, you can’t get them to do it. Or later when it’s something like potty training, it’s sort of like, okay, my kid doesn’t want to poop on the potty, I can’t make them.

So there are many moments like that in parenting, where you’re just like, I can’t make you fall asleep. I can’t close your eyes for you. And I’m working so hard. But there’s nothing I can do.

DUBNER: I’m curious what you think that says about like humans generally.

OSTER: I thought you were going to say what does that say about me? Nothing good.

DUBNER: Well, because considering how long our species has been around, and considering how well we do at passing along some types of information from generation to generation — math comes to mind; Euclid did a lot so that I don’t have to. But do you think as a species we’ve been not all that successful in passing along parenting information? Because it seems like every generation is newly flummoxed.

OSTER: Yeah, I think part of it is just like, infants are really hard. So particularly little babies — what are you going to say? Like, the baby cries a lot. And a lot of times, you can’t figure out what they want. And every baby is a little bit different. I mean, if you think about your parents giving this advice, it’s actually really hard to remember, particularly, this very early time, which is sort of out of your control. Also, nobody wants to listen to their mom when she’s like, “Oh, why don’t you try this? That’s what’s worked with you.” “Get out of my face.”

That was the Brown University economist Emily Oster; her new book is called Cribsheet.

*      *      *

Freakonomics Radio is produced by Stitcher and Dubner Productions. This episode was produced by Matt Frassica. Our staff also includes Alison Craiglow, Greg Rippin, Harry Huggins, Zack Lapinski, and Corinne Wallace. Our theme song is “Mr. Fortune,” by the Hitchhikers; all the other music was composed by Luis Guerra.You can subscribe to Freakonomics Radio on Apple Podcasts, Stitcher, or wherever you get your podcasts.

Read full Transcript


  • Emily Oster, author and Professor of Economics at Brown University.