Mainframe Brain Drain Revisited

Although efforts by mainframe boosters to recruit and train new technologists are bearing fruit, there are not yet enough newly-minted pros to offset concerns about brain drain in most shops.

Just three years ago, researcher Gartner Inc. put forward a particularly tendentious recommendation. A generation of technologists with great expertise on the mainframe was about to retire, it noted. One upshot, according to research vice president Dale Vecchio, was a looming mainframe brain drain that had the potential to "surpass Y2K as an unavoidable challenge" for enterprise application development groups (see http://esj.com/articles/2007/04/10/debunking-the-mainframe-brain-drain.aspx). In some very special cases, Vecchio concluded, it might make sense for an organization to transition away from Big Iron altogether.

At the same time, Vecchio was careful to concede that bread-and-butter mainframe applications -- particularly those based on COBOL -- weren't going anywhere anytime soon. "A legacy language such as COBOL has at least another decade of useful life, with few practical threats to even small and midsize businesses … on the horizon," wrote Vecchio in a Gartner report (Impact of Generational IT Skill Shift on Legacy Applications).

Three years later, few (or fewer) professionals seem to be talking about mainframe brain drain. This doesn't mean it isn't an issue, of course. Initiatives by IBM Corp., BMC Software Corp., CA Inc., Compuware Corp., and other mainframe boosters to recruit, train, or otherwise equip IT pros to manage or program for Big Iron platforms have borne fruit, but -- on the whole -- the specter of brain drain still haunts many shops.

Consider CA's most recent (August, 2009) mainframe market survey, in which fully 70 percent of respondents identified an aging mainframe workforce as a "critical pain point" in their respective organizations.

Less than half (40 percent) of respondents said they were in the process of recruiting or training a "new generation" of mainframe technologists.

At the same time, of course, almost two-thirds (61 percent) of respondents indicated that mainframe management and/or application development was becoming "less specialized" in their environments.

The outlook, then, seems to be something of a mixed bag. Yes, shops are still concerned about the demographics of a still-far-from-youthful mainframe workforce. Yes, efforts by Big Blue, BMC, CA, and others to recruit and train next-gen mainframe technologists do seem to be bearing fruit.

These efforts, however, don't seem to have produced enough newly-minted mainframe pros to offset concerns about brain drain in most shops.

At least one industry watcher thinks that's likely to change.

Training Up on Big Iron

Shortly after Gartner published its report, industry veteran Joe Clabby -- himself a Gartner alum -- took issue with some of its conclusions (see http://esj.com/articles/2007/04/10/debunking-the-mainframe-brain-drain.aspx).

Clabby recently revisited Gartner's prediction in a report published by his consultancy, Clabby Analytics. Yes, he concedes, there isn't exactly a surfeit of mainframe skills; conversely, he counters, neither is there an abundance of Java or Linux expertise.

"What headhunters [have told me] in essence was '[I]f you think there's a shortage of skills in the mainframe world … you ought to check out the skills shortage in Java and Linux,'" he writes, arguing that "a skills shortage across the entire IT industry" affects even "modern platforms" such as Java and Linux.

Moreover, Clabby continues, things are trending in favor of Big Iron, at least for training. Given time enough, he suggests, initiatives such as Big Blue's zNextGen effort could produce enough new mainframe talent to offset projected losses.

"What … college professors [have] told me is that mainframe training is on the rise -- and that the number of students enrolled in mainframe curricula has risen sharply," Clabby writes, citing a five-fold increase in the number of college or technical students enrolled in mainframe-oriented programs (from 10,000 in 2005 to 50,000 last year). What's more, he points out, more than 600 institutions now offer mainframe-oriented training; five years ago, he points out, that tally stood at just 213.