SURVEY THE MILLENNIUM BUG
Are you ready?
The consultancies were the first to try to work out who had done what, and at what cost. The Gartner Group surveys 15,000 companies around the world every three months and ranks countries and companies on a five-point scale from preliminary activity to fully compliant. Lou Marcoccio, the ebullient director of its Year 2000 effort, claims that a year ago, 50% of the worlds companies had not even begun work on the problem; now the share of laggards has fallen to 23%. The United States and Australia are in the lead, he says; Canada, Britain, Israel and Sweden are six months behind; France and Italy six to eight months behind; Japan a year; and Germany a year to 18 months. Many large companies and other bodies in Eastern Europe, India and much of Asia and South America have started only in the past three months.
Cap Gemini, another consultancy, compiles a Millennium Index based on interviews with 1,100 firms in 11 OECD countries. In April it published findings broadly comparable with Gartners, with America in the lead and France and Germany trailing Britain. But some of the results are baffling. For instance, the Gemini survey also suggested that 1% of companies in France and 74% of those in Germany would not complete their Year 2000 work on time. Such figures are hard to square with Geminis finding that Year 2000 work will account for 90% of annual IT budgets in Germany and only 20% in France. Does that mean the French are wildly underestimating their problems? Or that Germany faces the bigger task?
One trouble with measurement is that those who know most are not disinterested parties. Computer consultants have an incentive to say that not enough is being done, whereas the companies themselves want to play down the problem lest shareholders take fright. And their IT departments, notorious for failing to deliver on time or on budget, naturally assure them that everything is under control. One of the biggest problems I had with project managers, growls Bert Berende, who once bought computers for the American government, was that the project was always on time until it was late. A 100-day project was on time for 99 days.
Another problem is the absence of an agreed baseline. Do you just count lines of code, or do you devise a more complicated formula? With embedded systems, do you count individual chips, or the systems in which they are integrated, or components? Nor is there is an agreed measure of compliance; and if an organisation has not really begun to tackle its Year 2000 problems, it may not appreciate how far it is from the finishing line.
Sources of information are starting to build up. Merrill Lynch, a stockbroker, published a bumper special report in July, signalling the arrival of the Year 2000 as something that investors cared about. Some regulatory authorities, such as Americas Federal Reserve Board and Britains Ofwat, the regulator for the private water companies, are now scrutinising their industries for compliance. Several stock exchanges, including Americas Securities and Exchange Commission and the Australian and New Zealand stock exchanges, now insist on the disclosure of Year 2000 information.
As more and more companies ask their key business partners whether they will be ready on time, some are having to employ extra staff to deal with such inquiries. In a few cases, trade bodiessuch as the Air Transportation Association, which represents Americas airlines, and IATA, which does a similar job internationallyhave taken on the task centrally, and are building up a database of the replies. The British Bankers Association is devising a standard questionnaire.
All this activity has produced agreement on two points. First, survey after survey picks up the middle-sized as most at risk, be they countries, cities, hospitals or factories. Second, only a small minority of countries or companies will complete the task before the millennium. That makes it essential to set priorities and draw up contingency plans.
Medium-sized companies certainly face special problems. They cannot get away with just replacing their PC from the local shop, yet they lack the resources to tackle this sprawling managerial task. Working without an IT department, with customised systems developed by small software houses, with enough computer-controlled machines to be a headache and with quite a lot of links to other computer systems, such companies are likely to face particular disruption.
We think only one in four companies with under 500 employees is well down the track, says Gwynneth Flower, who runs Action 2000, a government-financed task force in Britain. In Korea, a government survey in March found that 34% of small and medium-sized companies planned to do nothing. The Australian Bankers Association reported, also in March, that only a quarter of small and medium-sized firms had taken action, most of it inadequate. More than 40% had no plans. Even in the United States, a survey in May by Wells Fargo, a bank, found that half the small-business owners who knew of the Year 2000 had no plans to act.
Muddled though middle-sized organisations clearly are, some of their big brothers could yet turn out to be in more of a mess. Bigger companies have (pro rata) a much bigger problem than smaller ones, and testing is particularly difficult for them, says Mr Guenier, who heads Taskforce 2000, a campaigning body. Most smaller businesses have still got time to fix the worst of it. A large business that has fallen behind does not.
As for governments, they have generally been slower than companies to measure their progress and set priorities. Only in the United States are Federal government agencies being monitored week by week, and repeatedly hauled before the two houses of Congress to testify. Senator Bob Bennett, a Republican from Utah who chairs a Senate committee on the Year 2000, has become a scourge of the unready. Congressional chivvying has revealed huge differences among federal agencies (see table 4). The Defence Department, plagued with lots of elderly mainframes, ingenious but arcane computer languages and embedded systems in almost every piece of equipment, has made disturbingly slow progress. The Federal Aviation Administration, with a batch of venerable IBM mainframes, began late and is still behind schedule.
On the other hand, the Social Security Administration has already started an elaborate programme of tests with all the other computer systems with which it interconnects. The SSA also wins pats on the back from Congress for its contingency plans. It has been working out how to continue to pay benefits even if telecoms or power break down.
All this emphasis on measuring progress may explain why the federal government seems to be even further ahead of most other governments than private-sector America is ahead of corporate rivals elsewhere. But American government also has by far the biggest problem. On the basis of (admittedly elderly and controversial) figures, public-administration employees in the United States are seven times as likely to have access to a PC or terminal as their counterparts in Britain, and almost 26 times as likely as public employees in Japan (see chart 5).
Beyond central government, progress is motley and mediocre. Among Japans municipalities, says MITI, the industry ministry, only 30% have even started work. In June a report by Britains Audit Commission found that fewer than one-third of local authorities and National Health Service bodies had even developed strategies for dealing with the problem, and fewer than 10% had begun to think seriously about contingency plans. Health care is a particular concern in many countries: perennially short of cash, increasingly dependent on information technology and on processes controlled by embedded systemsbut notoriously short of IT expertise, or managers who understand the issue. Wise citizens will avoid falling ill over New Year 1999.