Last summer, Michael Vu, a 40-year-old independent IT consultant, found himself in a wholly unexpected place midway through his career.
He’d signed a three-week contract to help a major U.S. retailer with an enterprise reporting project. The initial work was so successful that the project was extended. As a consequence, Vu was suddenly deep in the world of Cobol.
Yes, Cobol, the programming dinosaur that was last hot in the ’80s. Cobol, notorious for its overrich syntax and overlong code. That Cobol.
Although he’d never worked in Cobol before, Vu actually had wanted to learn for a while. In the midst of predictions of a massive retirement by baby boomers, Vu saw an opportunity. “I said to myself, even if only 0.1% of those baby boomers are Cobol developers, that would open up a huge market.”
As Vu’s work on the project proceeded, he realized that the retailer had 10 years of code living in Cobol. And the next phase of the project depended on that code.
So Vu, whose training and experience are in C and C++, jumped in and learned quickly. And he wound up with a skill that enhanced his strategic value to the organization. “I ended up moving from just being a regular coder with no idea of how the business runs to being someone they’re relying on to extract business knowledge from their code base,” he says. He now spends 30% of his time working in Cobol, and he expects that to stay the same or even increase.
For Vu, working in Cobol feels a bit like discovering a lost art. “The shocker for me was that Cobol is still heavily in use, even when my client is using the latest in enterprise Java, C+ and Visual Basic technologies,” Vu says.
I know it’s an old man’s game, [but] I like the position of being the younger individual in the market.”
Brian Vance, age 30,
Grange Insurance
What’s going on here? To paraphrase Mark Twain, the reports of Cobol’s death have been greatly exaggerated.
Some 75% of the world’s businesses data is still processed in Cobol, and about 90% of all financial transactions are in Cobol, according to Arunn Ramadoss, head of the academic connections program at Micro Focus International PLC, which provides software to help modernize Cobol applications.
Because of the massive installed base, it would be too expensive to try to replace all that code, he says. Instead, many companies are looking for ways to integrate Cobol with newer applications.
The experienced Cobol programmers who can best do that job, however, are dying, or at least retiring. In a 2007 Micro Focus survey of its customers, more than 75% of CIOs said they would need more Cobol programmers over the next five years, and 73% were already having a hard time finding trained Cobol professionals.
Aging out the market
“Without a doubt, it is a challenge to find a developer in Cobol who is not nearing retirement age,” says Dale Vecchio, research vice president of application development at Gartner Inc. In 2004, the last time Gartner tried to count Cobol programmers, the consultancy estimated that there were about 2 million of them worldwide and that the number was declining at 5% annually.
“Cobol will head downhill quickly over the next 10 years … as baby boomers retire and there is insufficient recharging of the population,” notes Vecchio.
As Vu’s experience shows, that may mean career opportunities for IT professionals willing to learn and work in Cobol. (How long those opportunities last, however, is a subject of debate. See “Cobol: Going, but when?” for details.)
We surveyed Cobol programmers and companies involved in the Cobol field and determined that the market these days supports two types of careers:
* An emerging role in which the programmer serves as a bridge between Cobol code and new applications. Such jobs require people who understand Cobol, the business rules and processes on which old Cobol programs are based, and more modern languages such as Java.
* A more traditional programming path, in which the employee maintains and fixes old Cobol code in addition to writing new code, also still in Cobol.
The Cobol liaison role can be an interesting career path, says Ramadoss. “Cobol doesn’t stop at Cobol,” he points out. “You can integrate it into any modern technology.”
With the emergence of service-oriented architectures, companies are able to more easily reuse their Cobol code, notes Nate Murphy, president of Nate Murphy International, an IT professional services firm.
The 66-year-old Murphy, who has decades of mainframe and Cobol experience, sees a resurgence in the value of Cobol because of the emergence of SOA and IBM’s Language Environment, which provides a common runtime environment for combining many different languages, including Cobol.
“Now you can extend and add subroutines for other Web-based features that you need,” he says. “All of a sudden you’ve got a valuable asset in these old Cobol programs, and you can extend them and expand their capability without writing new code.”
The other career path is the more traditional programming job — maintaining and fixing old code as well as writing new Cobol code. While some companies are now offshoring this type of Cobol work to places like India — especially the maintenance of old code — plenty of organizations want to keep a certain number of programmers in the U.S., especially if their jobs are key to keeping critical business systems up and running.
That’s the position that Stacy Watts, a 28-year-old senior developer at Nationwide Insurance in Des Moines, is in. She’s been writing Cobol code for about seven years, and last year the company offered her a chance to remotely oversee a team of programmers in India. Watts designs the program and then parcels out the coding work to the India-based programmers in addition to doing some of it herself.
Watts says she’s not worried that her job might be outsourced. Even with the offshore programmers, “We still don’t have enough people to get all the work done,” Watts says. What’s more, she views the opportunity to lead the India team as a step toward a management role.
Although Watts studied several programming languages at school, including Visual Basic, C and Java, she naturally gravitated to Cobol. “It was the mainframe that came easier to me,” she says. “It made more sense to me.”
Cobol programmers frequently cite job security as one of the attractions of their career choice. Brian Vance, a 30-year-old mainframe programmer at Grange Insurance in Columbus, Ohio, started at the company five years ago, maintaining and updating old Cobol code. Today, he’s developing new Cobol code as the insurance provider branches out into several new states.
I’m hanging on with two hands to my keyboard, and they are trying to pry me away.”
John Walczak, age 31, Sallie Mae
The youngest of about 20 Cobol programmers at the company, Vance foresees a stable and secure career. “I know it’s an old man’s game. I like the position of being the younger individual in the market,” he says. “You’re going to have people retiring and nobody to fill their shoes. So I think my job stability is about as good as it can get.”
John Walczak, a 31-year-old Cobol programmer at Sallie Mae Inc. in Indianapolis, also says he’s satisfied and secure in his work. When he graduated from Eastern Illinois University, Walczak wanted to work on Web-related projects. But Sallie Mae hired him to work on Cobol, promising that he’d be able to move around the company and do other things.
After a couple of years, he did indeed have an opportunity to move to a team that was developing a Web site. But to Walczak’s surprise, he didn’t like it. “I thought I’d be building Web pages and doing graphics. But that stuff is already prebuilt,” he says. Instead, he was building code “behind the scenes — doing a lot of Visual Basic and some .Net code.” He decided to go back to Cobol programming.
Now the company is trying to persuade Walczak to move into more of a liaison role. After working at Sallie Mae for more than eight years, Walczak understands how its systems work. “So they want me to use that knowledge to help with project development and with project design,” he says.
Problem is, Walczak’s not so sure he wants to make the move. “I love programming. I love to code,” he says. “I’m hanging on with two hands to my keyboard, and they are trying to pry me away. I don’t want to go.”
Cobol: Going, but when?
Most industry observers agree that a dose of Cobol training can help your career in the short term. But will Cobol be around long enough to get you to retirement age?
Companies involved in the Cobol market like to point to the statistics — such as that 75% of the world’s business data is still in Cobol — to prove that Cobol, and therefore Cobol jobs, will be around for years to come.
Dale Vecchio, a Gartner analyst, isn’t so sure.
“I’m seeing an increasing interest in organizations extricating themselves from IBM mainframes and Cobol,” says Vecchio. “It’s becoming increasingly accepted that they can get off the mainframe and move to Windows or Unix or Linux. I expect that to continue over the next five to seven years.”
In addition, large companies are increasingly replacing custom mainframe applications such as human resources or supply chain management — often written in Cobol — with packaged software from companies like Oracle, he notes.
Employment opps abound
Nevertheless, Cobol programming is still a useful skill for IT professionals to have. “The world doesn’t need 100,000 new Cobol programmers, but it does need several thousand new Cobol programmers,” says Drake Coker, chief technology officer for Cobol at Micro Focus International.
“There is a lot of work out there for people who know how to take a new system with new technology and marry it to an existing system,” he adds.
How to get Cobol into your toolbox is another matter. Fewer and fewer U.S. colleges and universities now offer Cobol training. In the past couple of years, both IBM and Micro Focus have launched initiatives to encourage universities to train more mainframe programmers. Through these programs, the companies provide schools with free technology and courseware.
Although these efforts might keep some Cobol courses going, Vecchio doesn’t think they will do much to prevent the dramatic decline of Cobol. The efforts, he says, “are too little, too late.”
