REDUCING BIAS IN A PERSONNEL ASSIGNMENT PROCESS VIA MULTIPLICATIVE UTILITY SOLUTION

In this paper we consider a type of bias which stems from the mathemat ical algorithm often used to determine an optimal match between two gr oups. We compare two different solution concepts for the matching assi gnment problem: the traditionally stable solution vs. a multiplicative utility approach that should avoid the bias. Simulation modeling lead s us to conclude that: (1) With respect to all sizes compared in our e xperiment, applicants and employers groups were always treated far mor e equally under the multiplicative utility approach than the stable ap proach. (2) When using the stable algorithm, the size of the bias is a ffected by the size of the problem (i.e., the larger the problem size is, the larger the performance discrepancy between two groups of parti cipants). Further analyses of the sensitivity of the finding to differ ent assumptions (e.g., correlation in the preference orderings of the two groups, or use of nonlinear conversion from ranks to utilities) al so resulted in a superior performance of the multiplicative utility al gorithm in terms of a more equitable outcome.