As mentioned above, when materials less dense than lead are used, the mass of a bullet particular shaped bullet goes down.  The ballistic coefficient is sectional density divided by form factor.   The only way reduce the drag of a less dense bullet is to improve its form factor.  This leads to attempts at highly optimized bullet profiles.   

That is more difficult than it appears because as a bullet travels, it is not pointed directly into the oncoming air.  Even a fully stabilized bullet travels at a slight angle to the oncoming air.  Before it "goes to sleep" or becomes fully stabilized, things are worse.  The bullets angle of attack to the oncoming airstream can vary widely as it goes through various oscillations.   The drag for many bullet profiles is increased significantly at the larger angles-of-attack that occur during these oscillations.   If I recall correctly, Emeric once stated that he has seen doppler radar numbers that indicate some bullets are still oscillating at 600m.  In other words, the drag measured from the flight paths of real bullets can be much higher than expected.  

When talking beter form factors in practice much of it is in the bullet tip. Note how many solids have extremely sharp tips.