A Regression-based Approach to Library Fund Allocation

Abstract

While nearly half of all academic libraries use formulas to allocate firm order funds on behalf of particular departments or subject areas, few have adopted systematic methods of selecting or weighting the variables. This paper reviews the literature on library fund allocation, then presents a statistically informed method of weighting and combining the variables in a fund allocation formula. The regression-based method of fund allocation uses current, historical, or hypothetical allocations to generate a formula that excludes the influence of non-relevant variables as well as the influence of arbitrary or non-systematic variations in funding. The resulting fund allocations are based on the principle of equity—the idea that departments with the same characteristics should receive the same allocations.