Where Are Out-of-Wedlock Births Highest?
An interesting image went around Twitter this morning. It’s a map of the share of births outside of wedlock in Europe.
The map comes from a really interesting paper published in 2015 in working paper form by the Max Planck Institute for Demographic Research.
The map shown is striking. Sharp discontinuities at national borders might reflect measurement differences, but they very well might reflect policy differences altering the economics of unmarried childbirth, or even cultural differences. They could also reflect historical or institutional differences. For example, we can clearly see the boundaries of East and West Germany. We can also clearly see the borders of interwar Poland and, more broadly, of Prussia. We can see the cultural split between northern Italy and the Kingdom of the Two Sicilies. All sorts of fun stuff.
But this got me curious. What would this map look like for the US? I’ve made one using ACS data for U.S. counties, which are a bit smaller than the units used in the above map but not much, and it looks like this:
One thing that stands out to me is that the U.S. map is way noisier. My suspicion is that the European map has had some smoothing applied to it across regions, which may, as it happens, be part of why regional and nation trends seem so pronounced. But again, since I don’t know much about the source, it’s hard to say.
But we can clean up our map a little bit. Let’s restrict our sample to include only counties where the margin of error on our estimate is below-average, so the counties where we have the best idea what’s really going on.
Well, now we’ve dropped so many counties that it’s hard to have any idea at all what’s happening!
Let’s fill in our blank counties from state data. That is, we’ll take total state-level births and unmarried births, subtract the births and unmarried births estimated for our lower-error counties, and assign the remaining unmarried birth rate to all the missing counties. This will create heighten the appearance of border discontinuities but, spoiler alert, I think country-level statistical pooling is happening in that European data as well.
Okay, so the south has higher unmarried childbirth, as do parts of the more rural northeast and west.
But still, surely we can do better?
Well, yes, we can do better. For each county, we can take the statewide average unmarried rate, and the county rate. If the statewide rate is outside the error band of the county rate, we’ll just take the county rate. If the statewide rate is within the error band of the county rate, then we’ll take a weighted average: the bigger the error rate, the more heavily the statewide average will be weighted. This is a way of saying that, when the county data is good, we’ll take it, but when it’s not good, we’ll push it to be more like the statewide average.
This is my preferred estimate of unmarried fertility. I think the average error at the county level vs. the “true” value is probably the lowest of my examples. And yet, this estimate will bias towards making states look like very important factors. While most true, this map may also be the most misleading. Compare this map to the first map I showed. Or, to make it easier, just flip between this page and this page. State boundaries are tons more visible on this smoothed map than on the unsmoothed ones, because I used state data as a smoothing factor.
So how about that European data? Did they smooth out local differences?
I don’t know. The paper doesn’t actually say. Its data section gives some underlying sources, but doesn’t actually discuss the kind of data being used: birth registries? Census data? Survey data? How is spatiality determined here?
Given that the church refers to 2007 and that very few European countries had Censuses on or about 2007, my assumption is that the data reflects surveys and birth registries. Surveys require substantial size or interpolation to estimate small regions; comparatively few European countries may have such surveys. Even the monumentally big American Community Survey struggles with small geographies.
That leaves birth registries. But birth registries may not reflect the actual spatiality of fertility: if registration is by place of birth, localities with hospitals will have higher fertility. Where a person births a child may differ, and indeed quite often does differ, substantially from where the mother and child reside.
I’m not sure exactly what’s going on, and I’m willing to believe that some degree of smoothing could be appropriate. But I’d also suggest that readers should not read too much into data that has been geographically smoothed out.
Now, the fun part of this paper isn’t actually about 2007 unmarried fertility. It’s about 1910 unmarried fertility. Here are the two maps side-by-side:
Make of this what you will. But one thing that jumps out to me is that some places have had fairly durable unmarried fertility trends, whereas some have not. The authors of the paper discuss possible reasons for this. But These two maps right here seem like a striking example of the importance of history to economics and demography. Yes, modern western Poland/Eastern Germany both have high unmarried childbirth, as does the border with Czechia. But those areas had high unmarried childbirth in 1910 too! In East Germany, it rose. In Czechia, it rose as well, but was kind of flat-ish in relative terms. In Poland, there was comparatively little nominal increase, and basically no relative increase. These places all experienced communism, so the “communism” explanation is hard. It’s actually very difficult to come up with a really cohesive story.
However, what we do see is at both periods, national borders often show striking differences. That might mean that policy matters. Or it might mean small differences in how governments collect data can have a big impact. Or both. My bias is towards “both,” that is, that international statistics are rarely readily comparable if collected by different bodies with non-identical practices.
Check out my Podcast about the history of American migration.
If you like this post and want to see more research like it, I’d love for you to share it on Twitter or Facebook. Or, just as valuable for me, you can click the recommend button at the bottom of the page. Thanks!
Follow me on Twitter to keep up with what I’m writing and reading. Follow my Medium Collection at In a State of Migration if you want updates when I write new posts. And if you’re writing about migration too, feel free to submit a post to the collection!
I’m a native of Wilmore, Kentucky, a graduate of Transylvania University, and also the George Washington University’s Elliott School. My real job is as an economist at USDA’s Foreign Agricultural Service, where I analyze and forecast cotton market conditions. I’m married to a kickass Kentucky woman named Ruth.
My posts are not endorsed by and do not in any way represent the opinions of the United States government or any branch, department, agency, or division of it. My writing represents exclusively my own opinions. I did not receive any financial support or remuneration from any party for this research.
