Demographic Change in Germany

While the French statistical institute INSEE publishes a very comprehensive list of names given to kids each year, the same data for other countries is hard to come by. In Germany, we have the strange situation that there is no official statistic, because there „is no legal basis“ for the collection of this data, but some guy collects an almost complete list of given names each year as some kind of private project.

Now, this guy doesn’t provide anything like a complete dataset – though he seems to fuel the yearly „most popular baby name“-articles that pop up everywhere. But since 2015 every year not only the 500 most popular baby names in Germany are published, but also the 10 most popular Turkish-Arabic names in Germany and their position on the overall list.

From the position of these names on the overall list, it is not immediately possible to derive an estimate of the percentage of kids with Turkish or Arabic names. Consider:
If the distribution of names is such that neighboring names on the list have almost the same frequency, the first name of a – say – 20% minority will pop up very far down the list, when this gentle slope has reached 0.2 of the frequency of the first name (assuming the same distribution for both majority and minority).
If, however, the distribution drops very quickly, for example as a power law, 1, 0.5, 0.33, 0.25, 0.20, … the first minority name would already appear at position 5 or 6.

In this post I will use the distribution of the French names to estimate the percentage of Turkish or Arabic names in Germany. This is sloppy because the French names do not comprise a single ethnic group any more, so I should probably remove non-French names for a better estimate. But also in that case the distributions of French names, of German names and of Turkish and Arabic names in Germany probably differ quite a bit, so let’s bite the bullet and keep in mind that these estimates are very rough.

Basically, what we will do, is look up the frequency of the names in the French distribution at the positions in the frequency ranking that are occupied by Arab/Turkish names in the German list. The sum of the frequencies will be compared to the sum of the frequencies of the first 10 (German) names. This gives us a straight-forward comparison of the frequency of Arab/Turkish names with the frequency of German names in Germany.

If we assume that both the German names and the Turkish-Arabic names are distributed just like the French names, we get the following estimates for the percentage of boys with Turkish-Arabic names:
2015 – 19.0%
2016 – 20.9%
2017 – 21.7%
2018 – 21.8%
2019 – 23.8%

If I assume the different French distributions of different recent years I get a standard error of around 0.7%, so which French distribution I choose doesn’t seem to impact the result too much.

However, the results for the girls differ significantly.
2015 – 19.3%
2016 – 18.9%
2017 – 19.2%
2018 – 18.6%
2019 – 20.0%

It might be the case that the overlap of Turkish and Arabic names is larger among boys and therefore the female names do not in the same way reflect the large influx of Arabs after 2015. This also opens up the possibility that even the estimates for the boys are underestimates, because Arabic names and Turkish names might be two largely separate distributions.

But given that the last official statistic gave slightly less than 10% muslim babies in 2004 and given that in France the percentage roughly doubled within 15 years, an estimate of 20-25% doesn’t seem absurd.