Why is the wrong volume being scored?

For some reason, when I inactivate multiple scattering the wrong volume is getting scored. I have a simple geometry consisting of a sandwich of 50 micron of Kapton followed by 5 micron of copper followed by 3 mm of a 70/30 mix of Ar/CO2. I am shooting 971 keV electrons with random directions into this sandwich and recording the energy deposited in the gas mixture. When multiple scattering is activated the distribution of deposited energies seems reasonable. But when I inactivate multiple scattering the simulation seems to be scoring the energy deposited in the copper, verified by changing the copper thickness and seeing that the distribution of the energy deposited was roughly proportional to the copper thickness.

I am including the detector construction and the stepping action files with this post in case I am doing something incorrectly. Also, in case it makes any difference, my Geant4 installation is version 10.7.1.

B1DetectorConstruction.cc (9.0 KB)
B1SteppingAction.cc (3.5 KB)