Emory Report

March 30, 1998

 Volume 50, No. 26

First Person

William Morse says computer
legacy systems must go

There is a crisis coming in technology. The New York Times recently reported there are an estimated 200,000 to 400,000 unfilled information technology jobs open in the United States today. Over the next nine years another million positions are expected to be created. This is an amazing 6 percent of all jobs expected to be generated in the United States during the period.

Currently the United States produces only 36,000 graduates with computer bachelor's degrees per year. In addition, the field gains 50,000 others who enter the field without a bachelor's degree or with a degree from abroad. This means fewer than 100,000 people enter the field per year. Worse yet, even with higher salaries, polls show students in high school and college are not entering information technology in large numbers because of its bad image and the lifestyle that technology people are seen to have.

While still a bit far away, the ramifications for information technology departments at Emory will be significant. A solution will involve the cooperation of technology departments and their customers, rethinking what has been asked of technology departments in the past and re-evaluating how they have been run.

To understand the full impact of the coming crisis, it is important to know where we are now. The rapid changes of the 1990s and inefficiencies inherent in technology departments-not only at Emory but everywhere-have left a bad taste in people's mouths. Technology has gotten the nasty reputation of being extremely expensive, with very little return for all of its costs.

How could this have happened? All of this new technology, after all, was supposed to increase efficiency. How could it be that the very departments intended to deliver this so-called "electronic revolution" could end up themselves with reputations for inefficiency?

The answer to these questions is actually very simple and can be expressed by the following equation: more obsolete systems + larger numbers of diverse systems = increased cost, poor service or, more likely, both.

Over the past 20 years computing has gone through a real revolution. Before 1980 computing was done almost exclusively on big, centralized mainframes, which helped control cost and standardize work production. It was simply too expensive to do anything else.

But everything changed when IBM introduced the first desktop computer. The IBM Personal Computer sold millions of units in its very first year, and in 1982 when the graphical Apple Macintosh was born, it too became a runaway success.

Because it was possible for ordinary workers to have computers at their desks for the very first time, machines soon became ubiquitous from boardrooms to secretarial workstations, and the performance gains through applications such as Lotus 1-2-3 and WordPerfect were very obvious and real.

But the revolution personal computers caused in information technology was just as significant. In addition to supporting the back-end "mainframe" servers that were and still are necessary, they had to support thousands of decentralized systems, and in the 1990s, computer networks created yet another revolution. Suddenly hundreds of individual units had to interact directly with central systems, becoming infinitely more complex to maintain.

Yet these facts alone do not explain how information technology became so expensive to maintain and got the reputation it has today. As computing technology rapidly developed and new systems were set up, obsolete units were left in place with all their maintenance costs and infrastructure. Like cars, the older computers get, the more expensive they are to maintain.

Why business left old systems in place can be boiled down to human nature: people simply do not like change. Both within technology units and with the users themselves, there was repeated resistance to having to face the fact that technology went through revolution after revolution that made everything before obsolete. As time passed, information technology units simply hired new people to do new projects while leaving old projects and the people running them in place.

After 20 years of this practice, there is now duplicative effort, spiraling costs and significant bloat. Many technology departments are so weighed down today with legacy systems, they cannot afford to move forward-the costs of the past are simply too expensive to allow it. Even if a department does try something new, it often lacks the resources to implement the project correctly. The result is a further stained reputation for information technology as new projects fail out of resource "starvation."

Of course, the other issue at work here is diversity. Personal computers by their very name imply individualism: each person with one of these new machines has personalized the system with their own choice of software, look and feel, operating system, etc. It was bad enough when support technicians basically had to "learn" each new desktop, which by their nature have to work together. The diversity already in place made the job almost impossible. Again the result was spiraling costs and a need to hire vast numbers of desktop support technicians.

And so technology no longer held a "competitive advantage." Instead it became to many people a necessary evil. The problems with diversity and obsolete systems were allowed to persist because no one was willing to push for the much-needed, yet painful, changes that would have to be made in order to deal with the serious crisis now approaching.

There simply are not enough technology people to fill all the open positions created by this spiral of personnel, and the problem is only going to get much worse. The end result will be substantially higher people costs within technology units and, worse yet, positions that will remain unfilled no matter what the salary offered. This crisis is coming, but there is something that can be done now.

Change will be painful, but today's systems are designed to allow a few people to maintain the desktops of many. In order to afford these new systems and make them work properly, older systems will have to be retired on a tight timetable and diversity must be controlled. This is a tough political issue to face today, but it is far better to face it now when solutions are possible than five years from now when a solution cannot be had at any price.

William Morse is director of Information Services at the Rollins School of Public Health.

Return to March 30, 1998 Contents Page