Abstract
A method of Kallio for improving bounds on the optimal value of a linear program calculated from an intermediate iteration is used to improve Zipkin's bounds for an aggregated linear program. Both theoretical and computational results are given, demonstrating the improvement due to these new bounds.