Robert Kuttners article on the causes of the Y2K problem (Costly Cure for Y2K Myopia January 4, 2000) is an interesting mix of fiction and ignorance. Not only does Mr. Kuttner make the absurd claim that his overstated costs of Y2K corrections should be laid at Microsofts door, but he repeats already discredited claims about the failure of markets in choosing technological products while demonstrating his ignorance of the workings of those theories.
Mr. Kuttner states that problems of path dependence were responsible for the Y2K problem. Theories of path dependence rely on coordination problemsconsumers, wishing to remain compatible with others, might choose a product they know to be inferior if they think that most other consumers are going to choose that inferior standard. In theory, this could happen, although in reality there is no evidence that it ever has (as demonstrated in my work with Stephen Margolis). Nevertheless, even the theory is inconsistent with Mr. Kuttners usage since firms would not build in Y2K problems just to be compatible with other firms. Misery might love company, but that is not sufficient reason to intentionally become miserable.
Instead, firms with large Y2K expenditures were basically upgrading and replacing their old COBOL programsmany of which were written prior to Microsofts very existencefor mainframes and minicomputers by programmers who never envisioned that their programs would still be running at the end of the century. The costs involved were going to be incurred sooner or later anyway, and therefore are not costs in the detrimental sense suggested by Mr. Kuttner, who chooses to ignore the benefits provided by these upgrades other than the Y2K fix.
It is also important to remember that programmers did not build in two digit years on a whimthey were trying to conserve what at the time was precious and expensive memory, allowing their programs to run faster and fit within the constraints under which they were working at the time. It may appear silly now to those unfamiliar with the history of computers, but it was not an irrational decision back then. It is possible that the users of the programs shouldnt have waited so long to upgrade, but the users fully absorbed the cost of their decisions. Thus, there is no externality, the second economic phenomena misapplied by Mr. Kuttner.
Although it is fashionable to blame Microsoft for many problems, only the most ardent Microsoft bashers would be so shameless as to go along with Mr. Kuttners outlandish claims that Microsoft is somehow responsible for the Y2K problem. Operating systems, such as Windows, do not run nuclear power plants, air traffic control systems, or hospitals - the examples cited by Mr. Kuttner. Specialized programs do, and that is where the Y2K problems arose. It is true that programs run on top of operating systems, but even here Mr. Kuttner is off base. The operating system that most of these old mainframe programs ran under was produced by IBM. Windows is barely ten years old so it could hardly be responsible for a problem which Mr. Kuttner himself admits had its origins forty years ago.
By way of analogy, imagine a town where many consumers continued to wear their shoes even when holes appeared in the soles, only wishing to change when rainwater would make their feet wet. After a long spell of dry weather, a rainstorm is seen heading toward this town and the populace rushes out to shoe stores and shoe repair facilities. The dollar expenditure on new shoes and shoe repairs is extremely high. Are these costs detrimental to society? Should the manufacturers of leather be blamed for these expenditures? Is there reason to believe that putting the government in charge of shoe purchases would improve welfare? A little thought reveals that the answer to these questions is no. Mr. Kuttner apparently believes that the answers would be yes, as long as Microsoft produced the leather.