In the late 1980s, I spent some time as a mainframe programmer at IBM. Conversations around the water cooler often had to do with some of the cryptic code written 2-3 decades earlier, buried deep within the system, that was incomprehensible to what anyone was writing at the time.

Now, 25 years later, the problem has grown exponentially worse. With a programming universe comprised of over 8,000 different languages, dated languages like Fortran, Jovial, and Cobol that lie buried inside corporate IT departments are coming back to haunt their host companies.

As an example, the day-to-day operations at the Mellon Bank of New York are based on 112,500 Cobol programs – 343 million lines of code – that run core-banking functions. Mellon Bank is not alone. Thousands of other companies have similar issues.

The ticking time bomb behind this problem is that the people familiar with this code are nearing retirement age. For companies that wait until that the institutional knowledge is gone, the costs for converting over may be as much as ten times higher than it would have been beforehand.

In an industry where speed is king, few college students want to learn the equivalent of programming Egyptian sundials when atomic clocks are driving the web. With a massive programmer shortage looming, it will come as no surprise that our newest venture at DaVinci Institute will be one to train next-generation programmers – DaVinci Coders.

History of Cobol

Cobol is one of the oldest programming languages, first appearing in 1959. Its name was derived from an acronym for COmmon Business-Oriented Language.

The COBOL specification was created by a committee of researchers from private industry, universities, and government during the second half of 1959. The specifications were to a great extent inspired by the FLOW-MATIC language invented by Grace Hopper, an inspirational visionary often referred to as “the mother of the COBOL language.”

Largely because of Grace Hopper’s influence, the percentage of female COBOL programmers is over 30% which is far more than most languages in this heavily dominated male arena. By contrast, only 6% of Ruby on Rails programmers are female.

More than 50 years after Cobol came on the scene, the language is alive and well in the world’s largest corporations, where it excels at executing large-scale batch and transaction processing operations on mainframes. The language is still popular because of its scalability, performance and mathematical accuracy.

Finding New Talent

A recent article in ComputerWorld does a good job of laying out the challenges ahead. New people entering the programming field couldn’t be bothered by the slow, tedious nature of programming in COBOL.

“College graduates with training in Cobol are in short supply. In Michigan, for example, state schools that offer Cobol programming have cancelled classes due to a lack of interest. “They can’t get anyone to enroll,” says Jonathan Miller, director of Saginaw County Information Systems and Services.”

“But some colleges are still providing Cobol training — with help from IBM. The mainframe vendor has developed curricula in association with more than 80 colleges and universities ranging from Brigham Young to Texas A&M. “We donate hardware and software, help with the curriculum, and they graduate hundreds of people every year,” says Kevin Stoodley, IBM fellow and chief technology officer.”

Having a gradating pool of “hundreds” hardly seems adequate for a language with a 50-year legacy and a massive bulge of talent ready to retire.

Salary Comparison

A recent survey by PayScale.com reveals some of the problem with recruiting new COBOL programmers. They simply don’t make as much money.


A senior software engineer in COBOL earns less than $80,000, while an experienced Ruby on Rails Developer can earn over $120,000 per year.


Currently the Ruby on Rails field is predominantly a male working environment comprised of 94% men and 6% women.

Even though there are an estimated 235,000 websites using Ruby on Rails, this is a young field. Only 3% have less than 1-year experience. 42% have 1-4 years experience, 30% have 5-9 years experience, and 25% have 10 or more years coding.

Many of the largest employers are using Ruby on Rails, including Amazon.com, Groupon, IBM, NASA, John Deere, Google, Living Social, Cisco, NASA, Oracle, JP Morgan, Twitter, Electronic Arts, New York Times, NBC, and many more.

What We Can Learn from the New York Stock Exchange

The anticipated retirement of in-house talent and the coming shortage Cobol programmers were a primary drivers behind NYSE Euronext’s decision to reengineer 1 million lines of Cobol on a mainframe that ran the stock exchange’s post-trade systems. While Cobol was dependable, it wasn’t viewed as maintainable in the long run.

According to ComputerWorld:

“Steven Hirsch, chief architect and chief data officer at NYSE Euronext, cites the need to make changes very rapidly as another key reason the stock exchange abandoned Cobol. “Ultimately, the code was not easily changeable in terms of what the business needed to move forward. We were pushing the envelope of what it took to scale the Cobol environment,” he said.”

In a recent survey, nearly half (49%) of survey respondents whose organizations don’t use Cobol say the reason is that the language is simply outdated.

The Coming Talent Wars

Several factors are converging that will make the decades ahead fertile territory for software engineers.

First, the Bureau of Labor Statistics is predicting a 22% increase in available tech jobs nationally by 2020. The demand for software services is already growing at a healthy clip.

Complicating these projections, many of the large corporations are beginning to contend with a retiring talent pool. Programmers well versed in the old languages, far more than just COBOL, are about to leave. In many cases, the cost of replacing the code after the resident talent leaves can be exponentially greater than when the in-house IT experts are still around.

Virtually every major industry will be conducting a code assessment over the next 2-3 years to determine whether the old code is worth saving. Every time a decision is made to “re-do everything,” either the company or its IT arm, will undergo a massive hiring surge.

In addition to everything else, mobile apps and mobile startups are becoming the new gold rush. A high percentage of existing programmers are ready to jump ship and start their own business when the conditions are right.

Young people today have demonstrated time and again that they are far more interested in launching their own business than they are in buying a house. And the business of choice invariably will involve a web operation or two.

Every startup in the tech world only increases the demand for additional coders. This becomes an incessant driver, one without enough talent in the pipeline.

The Genesis of DaVinci Coders

In looking over the opportunity landscape, we found a dearth of beginner-based training. People wanting to enter the programming field are left with the options of either going to a traditional college or learning on their own.

Since most people don’t have 4-5 years and $80-$100K to make the transition, or even 2 years and less money in the case of technical schools, traditional education is not a viable option.

On the other end of the spectrum, self-study programs that have recently become widely available and free, only appeal to the narrow spectrum of extremely self-motivated individuals.

Neither of these options does a good job of integrating students into the working life of coders by networking them into local companies or communities.

The best example of doing it right was Code Academy in Chicago started by Neal Sales-Griffin and Mike McGee in 2011. Starting their first class in August of last year, Code Academy focused in on people who are passionate and driven. With three times the number of applicants as to what they could handle, they found they had drilled into a deep and untapped opportunity.

Patterning our curriculum closely after what Code Academy is doing in Chicago, DaVinci Coders will be offering a full-immersion program based on an 11-week course with 10-hours per week of actual classroom instruction with homework and group projects filling virtually all of the non-classroom time.

Each class will be limited to 16 students. When they’re not in class, students will have their own pass code to use the adjacent coworking facilities inside the DaVinci Institute 24/7.

Students will each be assigned an industry mentor who will meet one-on-one with the students to answer questions and help integrate them into and familiarize them with the programming world at large.

The cost of this full-immersion program is $6,000. The first set of classes will begin on June 4, 2012.

Final Thoughts

With over 2,500 existing computer languages competing for mind share, we will naturally see the vast majority of them go away over the coming decade, with little more than a footnote in the tech history books to note their existence. However, the coding debris left behind will have to be dealt with in some fashion.

While the owners of this code see it as a problem, many others see it as a golden opportunity. Stale operating systems that were painful at best to make changes to, can now be rewritten in a vibrant language with interactive feature that allow them to move into the mobile spaces and social networking environments.

The number of coders and IT professionals will have to grow dramatically to meet the demand over the coming years. Already one of the highest paid professions in the country, programmer salaries are destined to climb much higher as skilled talent will be in short supply for decades to come.

By Futurist Thomas Frey

Author of “Communicating with the Future” – the book that changes everything

13 Responses to “The Coming Coder Wars”

Comments List

  1. James K. Poole

    Tom, Very interesting and timely article, but why do you and so may others insist on referring to "codes" and even "coders?" My experience with computer programs ls limited to a grad school class I took in 1964 and use of a few programs in processing research data, but I thought that computer programmers wrote computer programs in various computer languages, perhaps sometimes combinations of languages in subroutines. Even the positions listed in your article include programmers, not "coders." Using conventional English (a stretch for many computer geeks, I'll grant), "coding" would involve some form of encryption rather than use of an established, open language. Perhaps this could also apply to the use of proprietary programming languages not generally available to outsiders. So why a course for "coders" rather than programmers? Thanks for the article and your attention.
    • admin

      James, Thanks for your comments. The terms "programmer" and "coder" have been used interchangeably for a number of years. Regardless of the level of abstraction afforded by some of the higher level languages, what results is more computer code. Thomas Frey
  2. <a href='http://ngine21.com' rel='external nofollow' class='url'>Bill Hinkle</a>

    The best coders are those who can eliminate the most code, like 50-90%. I see coding as an essential art but a declining art. Nothing focuses the attention on coding like paying for it out of one's own pocket. Some technicians estimate the national code base at 5.3 Billion lines. This is a nation disgrace, a real testimony to inefficiency and competitiveness. Present coding techniques are nothing more than disasters looking for places to happen. Unfortunately, they are financed by well-meaning management who thinks they are doing the right thing. Management needs to guide (not follow) technology. The shoes are presently on the wrong feet. Bill Hinkle (303) 644-6321
  3. Spikosauropod

    It looks to me like there is a high demand for people who are expected to rewrite obsolete code and put themselves out of a job. I just don't see how you can make that career appealing. If I were a programmer, I would be more interested in the future of my own career than a flash in the pan opportunity to rectify the past. Am I missing something?
    • admin

      Scott, I don't think there's much danger of programmers being without a job anytime soon. Once they rewrite the outdated code, they'll be asked to add new features, additional capabilities, create a mobile app, and optimize performance. The only limit here is that of our own imagination, and there are plenty of creative people pushing the boundary of possibilities. Thomas Frey
  4. Spikosauropod

    Thomas, If I understand correctly, people who understand the old languages like Cobol are needed so that they can rewrite them into new languages. My point is that I would want to learn the new languages and would not bother with the old. Learning a computer language is a lot of work. I would want to devote all my energy to leaning the new ones. I can't see how you could inspire new programmers to learn the old languages, even if a demand for programmers who know both the old and the new are needed.
  5. <a href='http://be4one.com' rel='external nofollow' class='url'>Robert Lucore</a>

    A few thoughts: I learned COBOL in the Denver Public Schools in the early 1970s. I have long forgotten it, but it was not a difficult language. I would be quite easy for a programmer who is experienced in C, Java, PHP or Ruby to pick it up in almost no time. It’s just a little bit clunky because it originally tried to use English phrases instead of “math-like” equations for its statements. If there is a need for COBOL programmers, then it seems to me the market will respond pretty rapidly and rising demand will cause rising wages of COBOL programmers relative to Ruby-on-Rails or coders in other languages. It is hard to draw any conclusions from the two charts provided because the COBOL chart lists job titles and the Ruby chart lists years of experience. Also, those years of experience cannot be as Ruby-on-Rails coders, because Ruby-on-Rails was not even invented until 2005. So those coders started out in other languages. In summary, experienced coders can pick up another language relatively easily. If there is a need for COBOL skills, then employers need to pay a wage high enough to attract coders to start learning the language. Robert Lucore Student, Library and Information Science San Jose State University
  6. Rich K.

    I'm afraid this is a solution looking for a problem, but I get that the purpose of the article is to sell courses. When I rebuild an outdated website, the original is used as a template - what was working well vs. what wasn't. But the features are coded from scratch. And I won't even get into the wisdom of entering the field; I'll just say I suspect many of those programmers starting their own businesses are doing so because they're fleeing the industry. Programming can be a blast, but it can also be a living hell.
  7. <a href='http://www.cobolstyler.com' rel='external nofollow' class='url'>Ary Lemos</a>

    Thomas, Much of these problems could be solved if companies looked more closely at the software developed for managing mainframe cobol called cobolstyler professional. Everyone can get more information at the address www.cobolstyler.com. Ary Lemos The cobolstyler Team
  8. <a href='http://www.asapisystems.com' rel='external nofollow' class='url'>Kevin Weller</a>

    The problem with legacy COBOL code is not with its styling, but with its lower level of abstraction. It will take a great deal of work and redesign to convert the old functionality to newer languages and platforms. Those of us who have been in both worlds will be busy. :-)
  9. <a href='http://www.cobolstyler.com' rel='external nofollow' class='url'>Ary Lemos</a>

    CobolStyler does not intend to solve logic-programming problems, and does not discuss COBOL’s quality or adequacy. The fact is there are billions of COBOL lines being used right now and until other languages replace them, they must be maintained. CobolStyler offers a way to make maintenance easier. Again, the goal of CobolStyler Pro is not just making a style in Cobol programs with VSAM, CICS, SQL or DL/I, but especially to apply the rules of mainframe COBOL that are not always set correctly by programmers for lack of knowledge, correcting what is wrong, thus opening a window survival for the whole legacy, and more than that, creating facilities for programmers to do the maintenance and conversions to new language, if it is necessary. In summary, as the current lack on programmers with the specific knowledge of the mainframe COBOL language, CobolStyler Pro acts like a teacher correcting and pointing errors, facilitating the programmer's life in development & maintenance time, and improving the cpu usage in running day by day.

Leave a Reply