Calculating the cost of year-2000 compliance

Reviewer: Rudolph E. Hirsch

The benefits provided by this paper lie as much in its directions for undertaking global cost studies as they do in the help it provides for individual organizations in predicting the cost of their year 2000 (Y2K) conversions. The global aspects will be of interest to government and academic statisticians, for whom there is a detailed description of how cost surveys can be undertaken. The organization-oriented aspects, as the authors explicitly warn, must be taken with reservations, because the cost forecasts of the global studies done so far vary significantly. They all agree, however, that Y2K compliance is a large and expensive project. Some specifics are provided for organizational compliance. For instance, according to the authors, the two choices for affected software are to fix it or replace it. If the decision is made to fix it, the number of lines of code involved and the number of function points within that code will have a major influence on the cost of the project. Also, there should be little cost difference between fixing software logic only and expanding date fields to four digits. Y2K conversion cost estimates become more accurate as the conversion proceeds, but they also become larger: the final bill will probably be around 35 percent of the total IS budget until conversion is complete. There are some “soft” costs as well: the cost of forgoing opportunities while the IS staff copes with Y2K; and industry-specific costs, such as credit card issuers having to issue credit cards on a two-year cycle, not the usual three- or four-year cycle, because many card readers cannot accept the year “00.” The costs will be highest in information-intensive industries such as banking and insurance. The paper includes a short high-level checklist for organizations and makes one specific recommendation: If your Y2K conversion has not yet begun, it should, and immediately.