Steel + Long Reinf (n=n) vs. Steel + Long Reinf (n=3n) section property

In my composite steel girder run I notice the MOI in the negative bending region where the steel member and the reinforcing steel are the only parts contributing to the property that the 3N moment of inertia is different than the N moment of inertia.  I think N should be 1 since E of the girder steel is the same as E of the reinforcing steel (N = 29000/29000).  So why would the properties for Steel+Long Reinf (n=n) and Steel + Long Reinf (n=3n) be different?