Hmm. I’d probably just try not to think about that.

From The Data All Guilt-Ridden Parents Need by Emily Oster.

An interesting enough article and discussion about information and evidence as it pertains to some key family and child rearing questions. Oster is an economist and a good writer.

When my daughter, Penelope, was an infant, she was typically inconsolable between 5 and 8 p.m. I’d walk her up and down the hall, sometimes just crying (me crying, that is — obviously she was crying). I once did this in a hotel — up and down, up and down, Penelope screaming at the top of her lungs. I hope no one else was staying there. I tried everything — bouncing her more, bouncing her less, bouncing with swinging, bouncing with nursing (difficult). Nothing worked; she would eventually just exhaust herself.

I wondered whether this was normal. I’m an economist, someone who works with data. I wrote a book on using data to make better choices during pregnancy; it was natural for me to turn to the data again once the baby arrived.

And here, faced with crying, I found that the data was helpful. We often say babies are “colicky,” but researchers have an actual definition of colic (three hours of crying, more than three days a week, for more than three weeks) and some estimates of what share of babies fit this description (about 2 percent). But the same data can also tell us that many babies cry just a bit less than that, and almost 20 percent of parents report their baby “cries a lot.” So I was not alone. The data also told me the crying would get better, which it eventually did.

But I also found, more so than in pregnancy, that there are limits to the utility of general information. Parenting is full of decisions, nearly all of which can be agonized over. You can and should learn about the risks and benefits of your parenting choices, but in the end you have to also think about your family preferences — about what works for you.

She goes through a breast-feeding example. Is it good for the child? Sure, it has to be. Its natural isn't it? Well . . . That's where the devil is in the epistemic details. We no longer live on the plains of Africa. The fact that it is "natural" has less relevance given the actual environment in which we live.

The best point of the article is that even on topics which affect most people in reasonably consequential ways, we have little reliable and unassailable information with which to make informed decisions.

Let’s return from the land of magical breast milk to reality. Even in the most optimistic view about breast-feeding, the impact on I.Q. is small. Breast-feeding isn’t going to increase your child’s I.Q. by 20 points. How do we know? Because if it did, it would be really obvious in the data and in everyday life.

The question, really, is whether breast-feeding gives children some small leg up in intelligence. If you believe studies that just compare kids who are breast-fed to those who are not, you’ll find that it does. There is a clear correlation here — breast-fed kids do seem to have higher I.Q.s.

But again, this isn’t the same as saying that breast-feeding causes the higher I.Q. One study of Scandinavian 5-year-olds found that children who nursed longer had cognitive scores that were nearly 8 points higher on average. But their mothers were also richer, had more education and had higher I.Q. scores. Once the authors adjusted for even a few of these variables, the effects were much smaller.

In fact, the most compelling studies on this compare siblings, one of whom was breast-fed and the other not; these find no significant differences in I.Q. This same type of sibling study has also looked at obesity and, again, found little to no impact.

With complex, dynamic, evolving human systems (which definitely fits family formation and child rearing), useful information is hard to obtain, is often highly context specific, and often based on weaker than desired studies. We need to make decisions based on the available information but often the available information is of limited reliability and/or utility. That's when we fall back on folkways and heuristics. They may be crude but they frequently have utility.

Oster finishes with an example of pragmatism that breaks the ideal of fact-based decision-making.

I’m not trying to give advice. I’m just arguing that in many cases the data can be helpful. But if the data falls short and you still want advice, let me pass along something our pediatrician once told me. It was our 2-year-old’s checkup, and I had my usual list of neuroses.

“We are going on this vacation, and there are bees,” I said. “It’s kind of isolated. What if Penelope is stung? She’s never been stung before. What if she’s allergic? How will I get her to a doctor in time? Should I bring something to be prepared for this? Should we test her in advance? Do I need an EpiPen?”

In other words, I had built up this elaborate and incredibly unlikely scenario in my head. I needed someone to remind me that yes, this could happen. But so could a million other things. Parenting is not actually about planning for every possible disaster.

The doctor paused. And then she said, very calmly:

“Hmm. I’d probably just try not to think about that.”