>> Why would anyone want to work on an antiquated technology stack like Cobol?
Because it's actually a hell of a lot of fun. I did a year in a Cobol shop as a graduate developer at a large financial corporation. Working on a mainframe was the most fun I had on that job. And why not? You're logging on to a gigantic computer with millions of users and billions of transactions daily, with a text-based user interface that looks like it was designed by Tarn Adams. And I say that 100% as a compliment.
Seen another way, the "antiquated Cobol stack" is like a deep dive into the history of computer science and programming languages. You can see with your own eyes how stuff used to be done 50 years ago. And there is so much to learn. It's not just Cobol: there's a whole bunch of other languages like Rexx for example that is like javascript on a mainframe, or like, well, JCL which is an absolute horror to behold of course. There's a whole new operating system to learn, or three and there's a whole new computer architecture to become familiar with. How is that not absolute programmer heaven?
I mean, seriously, when I first got all the permissions and so on that I needed to work on a mainframe, I was giggling to myself like a little girl. "Really? They gave me access to all this?". It was like someone had given me the keys to the playground.
The job sure got a bit boring after a while, which is the reason I left, but for a few months it was just sheer tomfoolery, poking at things and finding how things worked.
>> The real solution is to think long term and rewrite all of these antiquated systems so that the next time there is an emergency it will be much simpler to find qualified developers.
That is not a sustainable solution. Fifty years from now people will be making jokes about "that antiquated Python stack" and state agencies will be ringing alarm bells for the lack of experienced young Python programmers. You can't just keep throwing out all the old code and replacing it with whatever new language is cool right now.