I was world-wide leader of Year 2000 services for Deloitte Consulting for 18 months to December 1997. In my experience, 75 per cent of all business IT systems need to be changed to avoid Year 2000 problems. About 50 per cent of departmental systems (such as stock control, laboratory data capture, packing and labelling) need to be changed. But the worst problem is that some 3 per cent to 5 per cent of "embedded" systems (controllers in electronic equipment) will fail in a serious way unless they are replaced.
There are billions of such systems in existence and, unfortunately, it is not generally possible to tell whether particular equipment contains a computer-based calendar and may fail. The clock may not be visible to the user of the equipment (as in some engine controllers). The clock may not even be used, but the equipment may fail its "power-up self-test" if the clock goes wrong. Equipment with identical model numbers may have different components, so you cannot rely on the results from a "representative" test. Manufacturers' assurances may be misleading or incorrect. Much equipment may be impossible to access, or difficult to test.
The business risks extend beyond individual systems and equipment. Will your suppliers fail? Will there be water, gas, electricity, telephones, transport?
Finding and correcting the problems is a skilled task. It needs to be done with greater professionalism than was used to build the systems or equipment the first time round. We are about to suffer from the consequences of believing that developing computer systems is a job that can be done reliably by poorly trained and poorly qualified staff. We will not solve these problems by creating an army of worse-trained and unqualified people and expecting them to make detailed changes on impossibly short timescales.
The money should be spent instead on training people to help draw up contingency plans, and on directing resources into the most vital areas of the national and international infrastructure.