Aggregation of constraints in integer programming

Constraint aggregation provides a method of formulating equivalent integer programs with a smaller number of constraints. This approach was widely researched in the seventies but its use was discounted due to large coefficients in the equivalent problem. We provide a method that yields numerically smaller constraint coefficients. This method has enabled us to investigate other computational issues relating to the use of constraint aggregation in solving integer programming problems, more thoroughly than has previously been possible.