The fact that many computer programs will be unable to recognised the year 2000 and think it is 1900 has been a profound worry for anyone involved in the IT world for the last three or four years. But it has been essentially a technical and financial issue: what needs to be fixed and how much will it cost. Companies and governments have been busy organising themselves to fix their computers, and see that their suppliers and customers were doing the same. But for the rest of us it has been, at best, an amusing quirk and at worst a bore. It has been hard to believe it matters. For example, someone must have mis-set the date on the little portable computer on which I am writing this column, and it says it is 22 February 1900. Silly - but not very important because I am not relying on the computer to tell me the date.
In the last few weeks, though, the tone of the stories has changed. Suddenly it is becoming an economic story. Many people now realise that whether or not the bug is the threat it has been cracked up to be, the fact that people are worried about the bug's potential destructive powers will mean that it will have an important impact on the world economy.
Thus banks are stocking up on cash. This may be partly because they fear electronic payments systems might fail but it is much more because they fear people will want a larger-than-usual stock of cash. Fund managers are planning to cut trading in the weeks before and after the millennium in case trading systems go down. Conference organisers are pushing forward conferences to November or back to March just in case air travel is disrupted.
So even if there were no direct impact - if every computer and every embedded chip in the world functioned perfectly - the bug would still lead to some disruption. What might this mean to the world economy? Well, in world financial circles perhaps the most alarmist view comes from Dr Ed Yardini, at Deutsche Securities in New York. He has upped his estimate of the chances of there being a "Y2K" global recession to 70 per cent. True, he assumes that there will be serious computer failures and he may or may not be right about this. But he puts the argument in an interesting way by comparing the shock from the millennium bug with the oil shocks of the Seventies. Then we were an energy-driven economy so any break in the supply of energy damaged the world economy. Now we are an information- driven one and we face the threat of a break in the supply of information.
The parallel is beguiling, but it seems to me that you don't need to accept his argument to believe that he may be correct in his conclusions. He may be right for the wrong reasons. This is because the bug is adding to economic activity ahead of the millennium and will subtract from it afterwards. Spending on IT investment, particularly in the US, has soared in the last three to four years and is now more than 6 per cent of GDP and more than half of all US capital investment. Naturally much of that investment would have taken place anyway and is associated with the extraordinary longevity of the American boom.
It is very hard to distinguish what spending is fixing problems and what is putting in new systems to improve productivity. So expect investment in IT to race on after the bug is a long-forgotten problem. But the very dependence on IT systems for the vibrancy of US economic activity is itself a concern if there is doubt about the vulnerability of the systems. And once the great burst of work now going on to fix the bug problem is over, it would be astounding if there were not some sort of pause in spending - if only to assess priorities for the future.
There are therefore likely to be two different types of post- millennial flop. One comes from fear: that through the summer and autumn both companies and individuals will stockpile ahead of the event, bringing forward activities (and hence purchases) they think might be affected by the bug. Then for a couple of months after the event (and maybe a month before) they will wait and see what is what. For industries that can stockpile their output this may not be too much of a problem. For those that cannot, like airlines, the prospect is alarming indeed.
The second type of flop will stem from a swing in the IT investment cycle. The investment cycle has long been recognised as one of the drivers of the more general business cycle, but until recently IT was not a large portion of investment. We have lots of experience of the disruption impact of other investment cycles: airlines buying aircraft in good times, only to have them delivered at just the moment when demand fell, or property companies starting to build office blocks in a boom, only to complete them and have them half-empty in the next slump. But we don't have any experience of an IT investment cycle. Maybe this will be the first one.
To this must be added the possibility of a third type of flop, a financial market flop. Markets go up and markets go down. You do not need to paint the more extreme horror scenarios to accept that present valuations, particularly on Wall Street, are stretched. If there is a general "the party's over" feeling around the millennium, it would be surprising if this were not reflected to some extent in share prices.
There is a precedent of sorts. In December 1899 the Dow fell by nearly a quarter. It subsequently recovered - but of course the US economy now is more vulnerable to market moves than it was then because a much larger proportion of the nation's wealth is tied up in shares.
We are a bit unlucky. The timing of the millennium comes at a bad point in the world economic cycle: towards the end of a long boom in America, but before Japan and East Asia has recovered from its most serious post-war slump and when the European economy is performing unevenly. In addition, the world is facing a tricky transition from inflation to deflation.
The oil shock came at a bad time too: a sudden rise in energy prices at just the moment the world was facing an inflationary boom. But just as we could not chose the timing of the first oil shock, we cannot chose the timing of the millennium - even if my computer thinks we have prematurely reached the end of the century and gone back to 1900.