In any case a 10% fluctuation can hardly be called "basically cut in half", as you put it. I am reading "10%" as "basically unchanged".

As it stands your argument is not even supported by the sources you are citing yourself.

Edit:

Ahah! When just comparing the single month of September we can find a decrease of about 50% though, according to the article (couldn't find a source for the number though, see below). So at least we have a number like that if we make the window so small that the data we have is beyond useful.

A 50% fluctuation on a monthly basis is perfectly normal though. For instance there was a +120% change in April 2018 / April 2019 (1517 vs 3345)[1] in total permits on a per-apartment basis. You can find plenty of such fluctuations though when just looking at months.

The data is pretty much completely useless on a month-by-month basis and can easily be skewed by a few large projects.

Also I can't figure out where that article is pulling its numbers from, because it quotes "more than 1700 permits for apartments last year in September", while the official statistics[1] just have 1301 when counting any kind of permit. For completeness, here is the press release referred to by the article, which also doesn't contain any number like that: https://www.statistik-berlin-brandenburg.de/pms/2019/19-11-0...

[1]: https://www.statistik-berlin-brandenburg.de/publikationen/St...