SURVEY   THE MILLENNIUM BUG

Countries that count
Further Reading:

Search archive

    Who’s the most millennium-compliant of them all?
 
AMONG many tales about the Year 2000 comes one from France. A British analyst, interviewing a main-board director of a leading French bank, asked what the company was doing about the millennium bug. “The Year 2000 question is a conspiracy cooked up by the Americans and the British to create a smokescreen and distract attention away from preparations for the single European currency,” declared the banker.

With the main exceptions of Britain, the Netherlands and Scandinavia, European countries have taken a strikingly relaxed attitude to the Year 2000. “The problems we expect are negligible and will not destabilise the economy,” averred Christian Perret of the French industry ministry in May. “Doomsday predictions are aimed to sell IT services; it is a formidable marketing tool.” Most European politicians have said nothing about the Year 2000. Even more than Americans, they belong to a generation that has little feel for IT. In Brussels, the issue has barely been discussed.

On the other hand, Europe’s politicians have said plenty about the euro, which has also been the main preoccupation of IT departments in banks and big companies in the 11 countries that will start to introduce the currency on January 1st next year. From then on, banks in those countries must convert all incoming payments into euro, and allow any client to carry out banking transactions in euro if he wishes. Institutions other than banks have three years in which to move across to book-keeping in euros (the issue of currency does not begin until 2002), but many large companies plan to begin at the start of 1999. That will rapidly force their suppliers and business customers to follow.

Accommodating the euro is at least as complicated as swatting millennium bugs. It requires not only extensive staff training, but new business processes, complicated new software and revamped computer operating systems to deal with the new currency symbol. (That creates special problems in Ireland, where the computer key which the Commission wants to represent the euro is already used for a Gaelic accent.) For European banks, estimates Merrill Lynch, the median cost of euro conversion is 2.6 times larger than the Year 2000 bill.

Will the experience of one be useful for the other? The danger is that software changes made by the team working on the euro will be disrupted by the team working on the Year 2000. The euro deadline, being a year earlier, has necessarily had priority. “We see the urgent competing with the important,” says Graham Brown of Neaman Bond, a consultancy that has been asking companies since autumn 1996 which problem they were concentrating on (see chart 7).


picture

Priority for the euro may explain the finding in several surveys that companies in continental Europe have generally done less to fix the millennium bug than their counterparts in the United States. “It is our impression that there is a striking difference in awareness and sense of urgency” between the United States and continental Europe, says a recent report by Goldman Sachs, an investment bank. One reason may be that the United States, having become more heavily computerised earlier, runs more of a risk of disruption. Europe, like Japan, has roughly half as many PCs per 100 people as the United States and—more important—proportionately about half as many mainframes.

But another explanation is simply that Europe began late. That is certainly the conclusion of the most detailed examination yet of Europe’s banks, published by Moody’s in May. Large banks, it reports, are well advanced. But Moody’s is worried about the ability of smaller banks (such as the savings and co-operative banks in Germany) to cope with the double challenge.

Should the
EU have postponed the introduction of the euro when the scale of the Year 2000 problem began to emerge? David Wright, a senior official in the Brussels office of Jacques Santer, the EU Commission’s president, scoffs at the idea. “Nobody in the Commission considered the link between the two,” he says. “The Year 2000 is still not seen as a serious issue here. It would have been politically unthinkable to have linked the two.” And when the idea was put to Sir Leon Brittan, the EU’s trade commissioner, his response was a merry laugh.

The Japanese enigma

Japan is also large but distracted. Outsiders regularly confess they are mystified by what is happening—or not happening—there. The Gartner Group puts Japan’s level of preparation roughly on a par with that of Mexico or Brazil. The published budgets of Japanese banks for coping with the Year 2000 are dwarfed by those of large banks in Europe or North America. Citicorp, for example, is spending $600m, whereas Tokyo-Mitsubishi Bank, Japan’s biggest, thinks that $100m will be enough.

Officials at the Bank of Japan insist that the country is taking the subject seriously. They cite three reasons for the banks’ lower spending:

•  Japanese banks still rely mainly on large mainframe computers, which are less expensive to survey and repair than the widely distributed client-server systems common in American banks.

•  Much remedial work was done in the late 1980s as part of a general overhaul of the country’s financial systems.

•  The people who introduced and overhauled Japanese IT systems are not—as in America—hired guns who stay for a year or so and then move off to another employer. Most are still on the same payroll. In western countries, programmers generally do not bother to write down what each line of code does and why—especially if they are not likely to be around to maintain it. By contrast, Japanese code tends to be extremely well documented.

Japan has other advantages. It has, for instance, an alternative way to write dates, using the emperor system (1988 was Showa 63; 1998 is Heisei 10). That can cause problems, because it is not always clear which system the two digits refer to. But, according to Takashi Ohashi, who works on the Year 2000 project at Fujitsu, a computer firm, small companies often find that the simplest way to adjust the dates on their applications programs is to switch to the emperor system.

Japanese financial institutions may also have spotted the problem earlier: most savings products run for five to ten years. And the Japanese Information System User Association (JISUA) had the foresight to propose back in 1992 that all programs written in Cobol (the most common language used in business systems) should adopt four-digit dates.

That said, Japan also has special reasons to worry. One is a serious shortage of programmers. In February the Japanese Information Service Industry Association (JISA), which represents computer vendors, estimated that some 2m-3m man-months of work remained to be done to get ready for the millennium, but Fujitsu reckoned at the time that the country had only 1.35m man-months of capacity. Even allowing for some hyperbole, that is an unnerving figure. In addition, neither regulatory pressure nor legal liability provide as strong a spur to action as they do in the United States.

Awareness of the problem is now rising sharply. Surveys by JISUA and JISA show that in the financial year to March 1997, 10% of businesses said they were aware of the problem; one year on, it was 60%. Officials from the industry ministry worry most about small firms and about embedded systems, on which Japanese companies have been slow to start work. “What the Japanese are spending is probably too small,” concludes Yoshinari Fujita of the Nomura Research Institute. “I don’t think survey responses are trustworthy. Respondents are still too naive about the size of the problem.” But then many Japanese companies are unsure whether they will survive until the new millennium at all.