Cookie's Chronicles: The Skeptical Reader: Do First-Born Boys Really Demand More Mommy Time?

I came across a parentdish.com article this morning that referred to new findings that women (in certain large countries - the US, the UK, Italy and Sweden) who have first-born boys are working less. Parentdish based it's story on a Wall Street Journal article - in fact lifted one entire section of text from the Wall Street Journal article and pasted it into their own. The Wall Street Journal references the abstract of a discussion paper by the Centre for Economic Policy Research (CEPR). My guess is that neither of these reporters read the full text of the discussion. I'd bet money that the Parentdish author didn't even bother to read the abstract and read only the Wall Street Journal article.

Here's the relevant text from the abstract of the CEPR discussion.

"We show that in the US, the UK, Italy and Sweden women whose first child is a boy are less likely to work in a typical week and work fewer hours than women with first-born girls. The puzzle is why women in these countries react in this way to the sex of their first child, which is chosen randomly by nature. We consider two explanations. As Dahl and Moretti (2008) we show that first-born boys positively affect the probability that a marriage survives, but differently from them and from the literature on developing countries, we show that after a first-born boy the probability that women have more children increases. In these advanced economies the negative impact on fertility deriving from the fact that fewer pregnancies are needed to get a boy is more than compensated by the positive effect on fertility deriving from the greater stability of marriages, which is neglected by studies that focus on married women only."

Somehow, from the above abstract, the Parentdish writer concluded that First-Born Boys Demand More Mommy Time. That's an interesting spin - it's the one that piqued my interest and sent me to their site to read the article - but it's not at all what the CEPR discussion paper found. The findings are simply that moms with first-born boys work less. I read the discussion paper and can not find a way to draw the conclusion that the Parentdish writer has. The CEPR discusses two potential reasons why moms with first-born boys work less: the "desire for a son effect" and the "divorce effect". Here is the text from the CEPR discussion paper:

"As shown by Bedard and Dech´enes (2004) the rate of marital dissolution is 4% higher for women whose ﬁrst-born child is a girl. We refer to this second channel as to the divorce effect. Since women in unstable marriages have fewer children over their lifetime, the gender of the ﬁrst-born child has ambiguous effects on fertility in countries where divorces are more likely. On the one hand, a ﬁrst-born boy increases the probability of marital stability (the “divorce effect”) and, as marital stability implies more births, it may also increase fertility. On the other hand, having a ﬁrst-born boy reduces the need of other pregnancies (the desire for a son effect)."

I find that the abstract doesn't represent the findings well enough and is confusing enough to be misleading. Perhaps this is why reporters are having a hard time getting it right. The conclusion of the discussion paper was well written and better reflects the findings of the study:

"We have shown that in the US, the UK, Italy and Sweden women whose ﬁrst child is a boy are less likely to work in a typical week and they do so for fewer hours than women with ﬁrst- born girls. Our estimates are statistically signiﬁcant and translate into quantitatively relevant labor income losses over the lifetime. The effect of the ﬁrst child sex is the combined result of at least two important sets of channels. To begin with, a ﬁrst-born son reduces fertility because fewer pregnancies are needed to have a son (the desire for a son effect). Because of lower fertility, mothers of ﬁrst-born sons should work more, and this is typically the evidence found in developing countries. But the sex of the ﬁrst child affects fertility also in an opposite way, by making the marriage more stable in case of a ﬁrst-born boy (the divorce effect). We show that in advanced economies this effect dominates and fertility increases when the ﬁrst child is a male. For this reason, in the countries that we consider, a ﬁrst-born boy decreases maternal labor supply."

This is why it's important to read not just the abstract but the entire article where possible.

While the numbers used to draw the above conclusions may be statistically significant, I'm not convinced that they are clinically relevant (i.e. meaningful in the real world). However, the CEPR did find a relationship between first-born boys and decreased maternal labour supply. I just can't get my head around the conclusion that first-born girls more often lead to divorce - 4% more often. Is this really a number we can trust? I'd want to see this study replicated, and I'd care more if Canada had been included in the study.

What's the moral of this story? Don't believe everything you read. In particular, when you read a report on a finding, see if you can dig up the original discussion or journal article from the people who researched the topic. Be wary of the spin that a reporter can put on a headline. The Parentdish article is very misleading. When you read an article that appears to take a number of quotes from another article and merely rearrange them (in this case in misleading ways), be skeptical.

Be your own judge. If you'd like to read the full text of the original CEPR discussion paper, you can download it here.

EDITED TO ADD: A friend in the research field has this to say about the study:

- First, I don't believe that gender has anything to do with working. What are the underlying variables? Did the researchers control for culture, or for predetermined decisions not to work after children? For example, maybe those women had previously decided not to work after children. Maybe those women are immersed in cultures where the mom is expected to stay home. I would have been interested to see the religious and ethnic demographics. I also think it would have been more interesting to see a longitudinal study where women were polled on their plans for work after children before they knew the gender.

- As for the higher divorce rate, again I think there are too many other variables which are likely to be underlying this phenomenon which were not controlled for. Were those marriages stable before kids? What beliefs do these couples hold about gender (i.e. Is producing a son expected or seen as an obligation)? What are the medical and behavioral histories of these children ( maybe those in the divorce category had children who experienced more issues)? What else happened after the baby was born (were there any other life problems such as financial issues, medical problems, family concerns)?

Essentially, I think that making the conclusions that these authors did is premature and misleading.

Thanks for your input, JR. These are excellent points!