Do I even have to argue this?

There are countless articles describing the problem that many COBOL developers are retiring and not enough new developers learn COBOL. 1 2 3 4 5 6 7 etc. etc. etc. I have seen a few articles that claim otherwise, but such articles are by far the minority.

If you are reading this, you are probably considering migrating away from COBOL yourself – and you wouldn’t consider this if you didn’t have a problem – which likely is coder shortage. You know your own pain better than me! When even the biggest COBOL-advocates like IBM, Micro Focus or COBOL-Works admit (at least between the lines) that there is a supply-shortage in the job-market for COBOL programmers – what is there left to say for me? I think the only thing I can do here is to outline how bad the situation really is and how bad it is going to become.

In the well-known Cobol brain drain survey, 55% of participants said that the ability to hire programmers is „worse or much worse“ for COBOL than for modern languages. Remember that this survey was from 2009! Back then, there were about 2 million COBOL developers left in the world. about 5% of all COBOL developers retire per year.

IBM tried to downplay the problem by saying that they had trained 180,000 people in COBOL. That probably puts your mind at ease, until you notice that this has been over the course of 12 years! that’s only 15,000 new developers per year on average! About 1.2 million COBOL developers have retired in the same time! 180,000 new COBOL programmers are a proverbial drop in the ocean.

Taking these numbers, we can conclude that there are only about 1 million COBOL developers left today. And by 2030, there should be only about 400,000 left. Worldwide. Remember that there are supposedly 220 billion lines of COBOL code still in use. Do you believe that each and every COBOL programmer can maintain 550,000 lines of code? Especially considering that most of this is mostly undocumented spaghetti-code?

Of course you can always hire programmers and train them in COBOL. I know a company that deliberately conceals from their job-applicants which programming language they use. But learning COBOL is much harder than many articles claim it was. Also: many employers report that their staff soon leaves for better-paid jobs after they have acquired the knowledge. So even if you teach people COBOL – they know that this knowledge is valuable – you still have to pay them royal salaries afterwards.

Salaries also reveal how bad the situation is: Senior COBOL programmers make around $90,000 per year on average. In Switzerland, it is even $120,000 per year. That is 50% – 100% more than programmers make on average. This couldn’t be the case if there wasn’t a considerable shortage in the COBOL job-market. 75-year old Bill Hinshaw, CEO of the „Cobol Cowboys“ says that he easily gets $100 per hour from his customers. Other sources speak of up to $2,250 per day. COBOL users are that desperate already! Imagine what it will be like in 2030! They would be foolish to work for less than $300,000 per year, $250 per hour or $5,625 per day.

Admittedly, people have predicted the death of COBOL for quite some time – but this time, the situation is different. We are not only talking about the language being ugly, unwieldy and far inferior to other languages. We are talking about the dying supply in an area that is indispensable to survival of the language. I honestly believe that this problem is the final nail in COBOLs coffin. And I am not alone in this. Notable experts agree with me – even down to the time-frame that COBOL has left. 1 2
Many, like Robert C. Martin or the O’Reilly media, consider it to be dead already.

Countless sources report that COBOL users are so desperate, they pull their old programmers back from retirement. 1 2 3 4 5 6 of course this only works as long as said programmers are still alive. Two articles 1 2 even mention a rumor about a company calling nursing homes, searching for COBOL programmers. I don’t think that’s more than a joke, though, to be honest.