2000: Year of Chicken Little?

Lots of money is being thrown at the "Year 2000 Problem," but everyone agrees on only one thing - it's impossible to tell what damage, if any, is possible.

Perhaps the newsgroup comp.software.year-2000 needs a name change, given the amount of apocalyptic squawking that goes on there. Something like alt.doomsayers.chickenlittle?

Participants in this corner of the Usenet community discuss what has become known as the "Year 2000 Problem," but is often abbreviated as "Y2K." At issue: Will the computer glitches expected to accompany the calendar flip from 1999 to 2000 wreak havoc, bringing down airliners, causing elevators to freeze, and deleting bank accounts?

The problem, which enjoyed crisis-of-the-month treatment on the cover of Newsweek last month, is by now fairly well-known. In the early days of computers, programmers used a two-digit shorthand to represent what in reality were four-digit years. While humans have little difficulty determining the century and millennium based on context (when you read the two-digit date atop this article, your mind automatically appended a "19," though a computer wouldn't have), the transition to the 21st century is expected to drive many calendar-reliant systems berserk.

How disastrous will the consequences be? "I think it's too complex a system to try to forecast," says Stephanie Moore, an analyst at Giga Information Group. Moore recently researched a Y2K bug affecting global positioning satellites, which will time out in August of 1999, potentially affecting navigation. "There are so many little problems happening all together that it could be really disastrous."

Other Y2K experts agree with Moore that predicting the scope of problems right now is impossible. Most companies and government agencies are still assessing the scope of their problems and investigating remedies. The estimates on how much will be spent range from the Gartner Group's oft-quoted price tag of US$600 billion worldwide to software productivity researcher Capers Jones' eye-popping figure of $3.6 trillion, which he says includes the costs of damages and litigation related to the bug. Howard Rubin, the chairman of Hunter College's computer science department and a leading Y2K researcher, says that between now and the end of the 20th century, fixing Y2K snafus will eat 25 cents out of every dollar spent on information technology.

Systems integrators, software companies, and consultants shift into salivary-gland overdrive at the mention of that kind of money. On 1 July, Gartner Group, the dominant force of the technology research arena, launched a new Year 2000 Strategies Service, staffed with 20 part- and full-time analysts. The bill? $20,000 a year, merely for advice on handling the problem. Keane Inc., one of the leading systems integrators focusing on Y2K issues, has so far inked deals with 230 corporations, ranging from $5 million to $100 million over several years.

And software companies like Data Integrity of Newton, Massachusetts, are churning out new tools to help organizations make the switch from two-digit dates to four. "You could see this as a gold mine," says Thomas Gary, a vice president at Data Integrity. "I look at it as though there's a fair opportunity to build a business." And a cash pile.

Peritus Software Services, a Massachusetts-based developer of Y2K software, had a stellar IPO earlier this month. Underwriters had priced the 3.5 million shares at $16, and the stock opened at $25.50.

Addressing the so-called millennium bug can be split into three phases: finding it, fixing it, and testing to make sure that the revamped systems will work properly. But then there's the issue of making sure that every supplier a company deals with has also taken steps to become "compliant," in Y2K lingo. "The prime example is a car company being totally dependent on suppliers of parts to manufacture cars," says Jeff Jinnett, an attorney at LeBoef, Lamb, Greene & MacRae in New York who concentrates on the legal ramifications of year 2000 compliance. "If they're missing one part, it'll shut down their assembly line."

Gartner Group estimates that 30 percent of organizations worldwide will experience failure of "mission critical" applications due to the date problem. "If just 1 percent of those go out of business, that's 0.3 percent of all businesses worldwide that will fail," says Gartner research director Matt Hotle. "You want to be sure that one of your major suppliers isn't one of them."

But will those failures produce the doomsday scenarios that make comp.software.year-2000 seem like it's moderated by Nostradamus? All signs point to "no." Hunter College's Rubin: "They'll be equal to the kind of things we're used to dealing with. The power goes out and air traffic control will go down for a while, or the stock exchange ticker will go out. I don't think we're going to see a new level of failures."

Companies that could be at risk seem to be taking the problem seriously, and doing quite a bit of rumor control as well. A spokesman for Otis Elevator says, "If you are in an elevator at midnight on December 31st, 1999, it won't change the operation of the elevator." But Otis is making fixes to a central computer system that tells its repairmen when elevators are scheduled for maintenance.

How big is the Y2K problem, in terms of actual work? Levi Strauss has identified 14,000 programs to fix, and has set a goal of patching 50 a day. Alamo Rent-A-Car is assessing 4,000 programs, composed of 9.6 million lines of code. And half of L. L. Bean's 200-person information-services staff is combing through more than 5 million lines of code to prevent the bug from interfering with flannel shirt deliveries for Christmas of 1999.

At Boeing, the company line is, "Boeing products worldwide will be Y2K-ready - not only those rolling out the doors, but those that have already rolled out the doors," according to public-relations manager Bob Jorgensen. Mary Jean Olsen, who heads one of the teams responsible for backing up Jorgensen's promise, says that some "minuscule" changes are being made to onboard flight-management systems, but stresses that "nothing would even have the potential to impact the safe flying of the aircraft."

But what about the FAA, and other government agencies that must wrangle with the Y2K problem on tighter budgets than Boeing? In a recent report on the federal government's efforts, the Office of Management and Budget stated that of the more than 7,000 systems considered "mission critical" to the operation of agencies like the Nuclear Regulatory Commission and the Social Security Administration, 59 percent are in need of repair, 9 percent must be replaced, 8 percent will be mothballed, 21 percent are already compliant, and 3 percent have yet to be evaluated. The OMB estimated that the Fed will shell out $2.8 billion for Y2K fixes, up from an estimate of $2.3 billion just three months earlier, but acknowledged that "more accurate estimates will become available as agencies complete the assessment phase." Gartner Group says Uncle Sam's bill will be a smidge higher than even $2.8 billion; last year it said the government would spend closer to $30 billion to fix the problem.

Some question whether Washington is taking the issue seriously enough. While Rubin received a letter from President Clinton stating, "I share your concern about the so-called Year 2000 Problem," the Hunter College prof believes Clinton and Vice President Gore need to speak out more. "Gore is supposed to be the techno-candidate," Rubin says, "and he declined to be interviewed by Newsweek about [Y2K]. This will affect the next presidential election, and it'll be more damaging for the current administration than the challengers. People will ask why Gore wasn't on top of it." And both Rubin and Y2K poster boy Peter de Jager agree that there's the potential for widespread social unrest if government benefits don't get paid out in the early months of 2000.

"Would you like to be in LA when they can't deliver the welfare checks?" asks de Jager, a Canadian technology consultant, whose ability to tap into the fervor this issue creates has put his name on an AMEX options index of Y2K companies. "There's a high likelihood of something like that happening."

De Jager sides with the denizens of comp.software.year-2000 on the potential seriousness of the millennium bug. In fact, he runs his own email list to help get the word out, in addition to delivering speeches around the world to help sound the alarm. While he says that the financial-services sector and the airline industry are making good progress on the problem, he is convinced that most companies and governments aren't doing enough. "It's real difficult to exaggerate what might happen, and the strange thing is that we're not doing anything to mitigate the risk," he says. "What do computers control, and what happens when they fail?"