The state of Schleswig-Holstein is consistently relying on open source

"Saves 15 million" on license costs, but how much will be wasted on the contractors involved, the lost productivity for state employees (especially the ones who depend on Excel, who will be converted too per the announcement)? And how much do you really save if you keep switching back and forth between M$ and Linux every decade, as state governments seem to enjoy doing?

They should switch to open-source for sovereignty. Not "cost". The fact that they mention "cost" as motivation and to secure buy-in is very worrisome. If you really want to switch to open source permanently and secure your sovereignty, you should invest more (making LibreOffice Calc as good as Excel? One can dream, but it's not cheap). Cost-savings show a lack of seriousness. How long until another government switches back?

How to know when they're serious: when the federal government hires an in-house team of (well-paid) programmers, and sysadmins. Not consultants. Put them in charge of public-facing and internal-use digital infrastructure, serving both the federal and state governments. Make them work to tailor a distro, or LibreOffice, to government needs. Invest in workforce training to keep their productivity up despite the switch.

And then, one day (let's dream for a second), that team could also pick new projects that serve the public interest, like a vulnerability research team (like Google Project Zero), or helping out with all those underfunded core pieces of digital infrastructure out there with only a single maintainer. Creating public goods is the point of a government.

It is better to spend 20milion on German contractors, then spending just 15m on licenses to foreign company.

At least the federal government loves to contract McKinsey, so a lot of the profit still ends up outside of the country. I didn't find any quickly accessible data on the state government in Schleswig-Holstein, though.