Thank you for pointing me to the document. The entire press release is a marvel of sophistry. From its first sentence ("Das Ergebnis der 59. Präsidentschaftswahl der Vereinigten Staaten am 3. November wird mit Spannung erwartet") to the later claim that "Seit 2017 ziehen mehr Personen von den USA nach Deutschland als umgekehrt", is the federal statistics office's communications office trying its best to claim that "Orange Man Bad so fewer Germans are moving to the US" without actually using Trump's name or, more importantly, with data actually supporting the claim. Most people are going to only read the headline and perhaps the first paragraph, and not see that the US hasn't been the most popular destination for German emigrants since 2004, so the recent trend is not a result of the Trump presidency. The US is still the third most popular destination as of 2019, only behind two German-speaking countries; it is ahead of every other country, including the UK and other countries that Germans can move to easily because of the EU and Schengen. Rather than Trump, it looks like the trend is a return to the pre-1990 norm. (More on this later.)
Regarding the chart you mentioned, the black line is presumably mostly Yugoslavian War refugees who used Germany as a stopover to get to the US (thus the late 1990s peak) and would not be applicable in showing native Germans' intent regarding moving to the US.
The red line does not show a recent trend toward "net migration from the US" (impling more Americans moving to Germany than German citizens moving to the US) but that, as the text says, recent population flows between Germany (German citizens) and the US have been almost even in gross terms. This is a reversion to the pre-1991 norm, as opposed to something unprecedented. Further, given that the US has about four times as many people as Germany, this implies that a random German citizen today is about four times more likely to move to the US than an American is to move to Germany.
Of course, these flows are not necessarily permanent; most are likely students, or workers on temporary assignments. Obtaining ciitzenship is a more meaningful indicator of interest. In 2019 4,745 German-born people obtained US citizenship while 1,205 Americans obtained German citizenship. This implies that a random German citizen today is 16 times more likely to obtain US citizenship than an American is to obtain German citizenship.
Further, 75% of those Americans obtaining German citizenship in 2019 do not live in Germany but are taking advantage of the right of return (encouraged, from 2006, by the German Citizenship Project (https://en.wikipedia.org/wiki/German_Citizenship_Project)), while the US does not have a similar law. Only about 300 of Americans obtaining German citizenship actually lived in Germany. This implies that a random German citizen today is 48 times more likely to obtain US citizenship and to move to the US, than an American is to obtain German citizenship.
>I'd love to look at your data, but that migration DRC link points to an apparently empty wordpress instance.
Sorry; didn't realize that the domain had expired. The version 4 I mentioned of the dataset is now at http://www.sussex.ac.uk/Units/SCMR/drc/research/typesofmigra....