Caffeine, Alcohol, and Television: What’s True?

Conventional wisdom—and most official recommendations—tell us that we must abstain entirely from coffee and alcohol for the duration of pregnancy to avoid very serious consequences. And once our kids are born, experts urge us to keep our children away from screens to prevent behavior problems and poor academic performance in the future. But what does the data say?

Is it really the end of the world to have a cup of coffee in the morning? How about an occasional glass of wine? What is wrong with letting our kiddos watch an episode or two of Dinosaur Train?

When economist Emily Oster got pregnant, she also got curious about the advice she was getting. Some recommendations were based on her age alone, and sometimes she found it difficult to get any answers at all. So, she started digging into the data.

Eventually, Emily’s research became Expecting Better, and today, she joins me to discuss her approach to writing for a popular audience and the somewhat unintentional way the book was born. Emily describes how her background as an economist shapes her decision-making and how to go about judging the quality of a particular study. I ask Emily for insight on the data around caffeine, alcohol, and screen time for kids, and she explains why two people can see and understand the same data—yet make different, and equally valid, decisions. Listen in for Emily’s take on making corrections to our thinking based on new evidence and get a sneak peek at what’s covered in her forthcoming book!

The Startup Pregnant Podcast Episode #080

Some quotes from the episode 

  • “The most disruptive thing work-wise is just that a lot of the time I am also trying to think about some aspect of parenting.”
  • “I like the idea of trying to help people understand why economics and data is useful in making decisions about your life.”
  • “Two people can see the same exact data, can understand it exactly the same way, and could still decide to do something different—and those could both be the right decision.”
  • “When you’re reading some new fact, you want to [ask yourself], ‘Does it seem likely this is true?’”
  • “If you want to have two cups of coffee … a day, there’s really no evidence that that’s a problem.”
  • “It’s just really challenging to help people see the nuance in the data and say, ‘There isn’t really a yes/no answer to this.’”
  • “Drinking a lot of alcohol in pregnancy is very dangerous. It can lead to pretty significant birth defects.”
  • “If you want to occasionally have a drink, the data suggests that that does not have negative impacts on your kid.”
  • “What is the right approach to policy? It is an open question whether telling people not to drink at all is a productive way to avoid the kind of high-risk drinking that leads to serious consequences for kids.”
  • “It’s not appropriate to infantilize people and infantilize women and say, ‘Well, I’m not going to tell you what the truth is. I’m just going to tell you what I’ve decided is the decision you’re going to make.’ I don’t think that’s okay.
  • “It was, of course, super-embarrassing because I was this junior researcher, and this was by far the most famous thing that I had written—and it was just not correct.”


Emily Oster is a highly-respected economist and professor at Brown University. Her research interests range from development and health economics to research design and experimental methodology. Emily’s work has appeared in the Wall Street Journal, the bestseller SuperFreakonomics, and FiveThirtyEight, among many other publications, and she is the author of Expecting Better: Why Conventional Pregnancy Wisdom is Wrong—and What You Really Need to Know.



Thank you to the sponsor of this episode: Aeroflow Breastpumps. They are dedicated to making the hassle of getting your breast pump a little bit easier—actually, a lot easier! Head to to have them help you qualify for a free breast pump through insurance.

All of our sponsor offers are available on our website for you to grab the perks and discounts offered to podcast listeners: