In real world shootings, the 32acp and the 380acp are virtually equal in effectiveness. In comparing both cartridges in real world shootings where only one round was fired and a solid torso hit was recorded, the best 32 round (a 60gr JHP) stopped the attack 63% of the time with 83 actual shootings studied. The best 380 round (90gr JHP) managed to stop the attack 69%-70% of the time out of 78 actual shootings studied. Worst rounds in both the 32 and the 380 were both FMJ rounds and the percentages were 50% and 51%, respectively. Just for comparison the best round out of a 2" 38spl managed 67% (LHP+P) and the worst managed 49% (LRN).
The many variables involved in real world shootings make it impossible to predict with any certainty exactly how effective any round will be in stopping an attack (i.e. you CANNOT say with certainty that these rounds will stop an attack at least 63% of the time with one shot). However, the comparison above is a good "apples to apples" comparison of the different calibers and bullet types in defnesive shootings (i.e you CAN say with authority that they are very close in effectiveness to each other, and the best JHP's are significantly better than the best FMJ's regardless of the caliber.)
Based on the fact that all 3 are basically equal in real world effectiveness, my pocket carry gun is a 32acp because:
1. It is smaller than the 380 or 38.
2. It recoils less than the 380 or 38, which means I can fire it faster. Since it seems conclusive that any of the three will take more than one round to stop an attack a lot of the time, that is important.
3. I carry the fastest JHPs that are reliable in my gun since they seem to be universally more effective than FMJ's in real world defense shooting, no matter the caliber.