Hold up! Marriage and fertility are still related

Robert VerBruggen has a thought-provoking post at RealClearPolicy today about the relationship between marriage and fertility internationally, or at least in large economies (it’s fairly rare to see demography stuff, so this is great). Playing off an earlier Phillip Cohen post on inequality and marriage (which is very good), the two variables he examines in his scatter plot are the Total Fertility Rate and the percent of adults who are married – at least I’m assuming this is the variable, given my eyeballing of his scatterplot. VerBruggen finds that there’s no relationship between the percent of adults married and the TFR, and concludes, “I find it fascinating that two phenomena as tightly bound to each other as marriage and childbearing aren’t correlated with each other internationally.”

This may be the case – and I think the wider point he’s making, that marriage is no longer a synonym for family formation in the developed world, is an excellent one that deserves inquiry – but that particular choice of variable to represent marriage isn’t great. It’s skewed by a number of factors, including differential mortality and median age.

For example, Russia’s relatively low proportion of adults who were married is almost certainly due to the massive gap between male and female life expectancy. At birth, the gap is 12 years. For comparison, the gender mortality gap in the United States is 5 years, which is right around the average for all countries in this analysis. This wide gap is due to a number of factors including smoking, alcoholism, and the gutting of the public health system during the era of rapid privatization in the early 1990s. Russia’s adult population is chock full of widowed adult women, so it’s no surprise they score relatively low on that variable.

Mean age at first marriage is a much better measure to determine if there’s a relationship between marriage and fertility in the industrialized world. As the correlation table below demonstrates, there is a demonstrable link between the ages at which people get married, the age at which they have their first child, and the number of children they ultimately have (fertility rate). This certainly makes logical sense. In other words, while more marriage on a population level may not necessarily mean more babies, earlier marriage usually does.

Perhaps more interesting is the link between marriage, childbearing, and certain expressions of policy and gender equality, such as female labor force participation, average weeks of maternity leave, and GDP per capita. Unsurprisingly, the Nordic countries – wealthy, with highly developed welfare states and low inequality between genders– have among the oldest ages at first marriage and childbearing, and the lowest total fertility rates.

(all bold correlation coefficients are significant at p<.05, those with two asterisks are significant at p<.01; apologies for the formatting)

My data is available here (and a deep thank you to VerBruggen for providing his data).

The post Hold up! Marriage and fertility are still related appeared first on Conor F. McGovern.