No One Here Can Do The Work
First, look at reality. There are no new COBOL programmers coming out of American universities. The least sexy area of study in the computer science department is COBOL—if you can study it. The name even makes you want to yawn—Common Business Oriented Language. The people that know COBOL started in business fresh out of college in the 70s. Simple math puts the younger ones close to their sixties.
Outsourcing Core Business
The risk? Knowing the language is only a small fraction of the knowledge required to maintain a system. Losing the understanding of how the few million lines of code run, the subtleties of how they support the business, and interaction of the modules is a huge risk. No amount of documentation will ever replace the experience that came from changing one line off code that left of a municipal tax. It was not enough to make the invoice run look real low, but three weeks later when taxes were due, someone in accounting noticed something very wrong.
The Politically Incorrect Option
One option, if you cannot find trained students, is to bring the COBOL developers onshore. The expense is higher, but the technical risk much lower. The real problem is the massive political undertaking. Headlines reading "Unemployment At Record Highs, Company Hires People From India," would be a public relations nightmare. Politicians would spin it this way and that to make it beneficial them or detrimental to their opponent.
Management has been painted into a corner. Unfortunately, they are holding the paintbrush. The solution is never getting to the point of obsolescence.
Averting The Situation
We all know the answer. When we buy a house, we know we have to do maintenance. Small items, like painting and replacing carpets, are done on an as needed basis, and can be postponed when budgets are tight. Replacing the roof or rotting LP siding are projects that we think about years in advance—before the wind, the rain, and the bugs encroach. To protect our investment, we modernize the entire house, adding a garage, or remodeling to add another room or two. If it looks like the neighborhood is heading into disrepair, we sell as quickly as possible and move to a better investment—one that will take us into the future securely.
Unfortunately, businesses that are stuck with difficult to maintain, twenty-year-old systems only have themselves to blame. They chose to try to get one, two, three, or more years out of that "roof" and now the upgrade is getting to be impossible, since no one works on slate roofs anymore. Averting this problem needed impetus a decade or two ago. Taking the expense and challenge to upgrade systems, migrate to newer technology, and stay in the mainstream instead of letting technical debt increase incessantly.
The winners are the people in India, China, or Vietnam that are getting new jobs. The losers are corporate America that has just found another way to kick the can down the road another few years.
Conclusion
The lure of postponing system replacements for short-term savings and improved quarterly reports only delays the inevitable. These projects are tough; they are laden with minefields and booby traps. The only path to success is a superlative team, strong planning, and contingency funds. When it is all in place, plan a little more. But, it goes nowhere without the team. This requires leadership and continuous training of your staff in the technology that aligns with your company's strategic plan.
(You do understand how technology relates to the strategic plan, right? Okay, you had better spend a few days digesting the strategic plan, getting your team together, and building a technology roadmap.)