A voice coding rate is selected for a voice call involving a user equipment (UE) device based on an air interface efficiency of the base station serving the UE device. The air interface efficiency of the base station is determined based on at least one of (i) a beamforming capability of the base station, (ii) a multi-user multiple-input multiple-output (MU-MIMO) capability of the base station, or (iii) an antenna configuration of the base station. The voice coding rate could be selected by either the UE device or by the base station. The UE device transmits to the base station during the voice call one or more voice frames that convey voice data coded at the selected voice coding rate. During the voice call, a new air interface efficiency may be determined, and a new voice coding rate may be selected based on the new air interface efficiency.