Effects of cataloging volume at the library of congress on the total cataloging costs of American research libraries

How should the library community organize its efforts to keep up with a cataloging volume that grows at about 5% per year? The decision is modeled as the problem of selecting the optimal value of a single variable, s, which is the fraction of all cataloging done centrally. It is assumed that the library community has B titles to catalog, and that the G (where G=s·B) most widely acquired titles will be cataloged centrally (e.g., at the Library of Congress), the cataloging records then being distributed to all libraries that have acquired each of the G titles. The other variables used in the model are the distribution of “popularity” among titles and the costs of cataloging books centrally or locally. The optimal s value is first determined in the general case. Then numerical estimates of the present situation are made, which indicate that s=100% is optimal now, meaning that everything (820 titles per day) should be cataloged centrally. At that rate, it is estimated that all cataloging could be done for $0.88 million per day, compared to the $1.07 million that is now being spent with s=40%. It is suggested that this potential savings of over $0.1 million per day, due to decreased duplication of effort by local libraries, could easily justify the use of a computer system to allow the value of s=100% to be reached.