THE BAD NEWS : ALL HANDGUN CARTRIDGES HAVE PRETTY CRUMMY BALLISTICS 

The truth is that ALL semiauto handgun cartridges boast relatively pitiful ballistics.  If you want to make certain you kill your attacker, pack a hunting rifle.  If you're carrying a pistol, you're carrying a compromise.

THE GOOD NEWS : BALLISTICS AREN'T THAT IMPORTANT

Contrary to popular belief, the ballistics of the cartridge in your pistol are NOT the most important factor controlling whether your pistol makes you safer.  WHETHER you have a pistol with you when you need one is infinitely more important than what CALIBER that pistol is.  You may have your big-bore pistol packed with the most lethal pistol cartridges in the world, but it will do you no good whatsoever if it's locked in your gun safe when you need it.  Similarly, a low-powered bullet that HITS an assailant will do infinitely more damage than a high-powered bullet that DOESN'T.  A .22 bullet hitting an attacker between the eyes will have much more effect than a .44 Magnum bullet glancing an ear.  Two hours at the gun range improving your aim will generally make you a more lethal shooter than two hours at the gun show buying a larger-caliber gun.

IS THERE A 'MINIMUM' PENETRATION?  A 'MINIMUM' EXPANSION?

There are a whole lot of "rules of thumb" regarding the "minimum acceptable" penetration, etc., but they are only that.  It will be the rare bullet that fails stop an assailant only because it lacks that extra 0.5" of penetration, or that 0.05" of expansion.  Could there be a situation in which a bullet would've stopped an attack had it moved only 0.25" further?  Sure.  Is it a likely situation?  No.

This isn't to say, of course, that there's little difference in tissue damage between a bullet which expanded to 0.8" and a bullet which expanded to 0.4", or that there would be little difference between a bullet which traveled 4" through vital tissue and a bullet which traveled 10" through vital tissue.  There obviously would be a substantial difference in tissue damage.  The point here is that, while LARGE differences in ballistic behavior will (all other things being equal) translate to LARGE differences in the level of tissue damage, SMALL differences in ballistic behavior will (all other things being equal) translate to SMALL differences in the level of tissue damage.

Keeping in mind the above idea, take a look at the following chart:

This is an X-Y plot of the ballistic performance of various cartridges, with penetration plotted along the X-axis and expansion plotted along the Y-axis.  The penetration numbers range from about 6 inches to about 23 inches, with an average penetration somewhere in the range of 12-14 inches.  The expansion numbers range from about 0.2 inches to about 0.8 inches, with an average expansion somewhere in the range of 0.50".  In general, more expansion is better and represents more tissue damage.  Penetration is more complicated.  While adequate penetration is important, overpenetration is generally considered undesirable, as it represents wasted energy and increased risk to innocent bystanders.

As you might have guessed, the data points on the graph above came from a variety of different cartridges.  In fact, the above graph includes a hodge-podge of .22 long rifle, .32 acp, .380 acp, 9mm Luger and .45 acp.

The graph below shows the data points identified by caliber:

As you might expect, the results showing the highest penetration and expansion numbers were generally .45s.  That said, some of the better 9mm cartridges outperformed some of the weaker .45s.  The 9mm and  .38 special are occupying essentially the same real estate.  Along a similar line, the better .380s are competitive with the weaker 9mm cartridges.  The .32s are, as a group, just a notch below the .380s--but not by that much, really.  The .22 long rife brings up the tail end of the group.  By the time you get down to the .22 options, you're looking at cartridges that offer 50% of the penetration AND 50% of the expansion of a good 9mm.  That is, of course, a substantial drop in performance.   Those folks who consider the .22 just too weak to be a good self defense option may have a point.

In some ways, the above chart tells us what we already knew--namely, that a .45 has more gusto than a 9mm, which has more gusto than a .380, etc.  On the other hand, this chart illustrates a point that some of us may not have realized--namely, that the identifiable differences in the ballistic performance of these cartridges, while significant, are not quite as dramatic as some would have you believe.

Anyone who'd completely discount the .32 acp should consider the substantial number of police and military organizations, including the German Wehrmacht, who found the .32 acp sufficiently powerful for their purposes.  A .32 pistol was, after all, powerful enough to start World War I when it killed Archduke Franz Ferdinand.  Another .32 pistol was powerful enough to end the fighting in Europe when it killed Adolf Hitler.

Even the lowly .22 is far from useless or harmless.   The .22 Long Rifle has been used by the Israeli Mossad and is in current use by the U.S. Navy SEALs.  The British government issued .22 pistols to their special agents in WWII after testing the effectiveness of the .22 Long Rifle against other available pistol cartridges. Chechen snipers were known to use the .22 Long Rifle very effectively for sniper duty against Russian soldiers.  Anyone who claims that a .22 bullet just won't do much damage should tell that to this turkey. 

Bottom line: the .22 Long Rifle and .32 acp cartridges can, and do, kill--but there are certainly other options.

OK, BUT WILL "CARTRIDGE X" STOP THE BAD GUY?

Probably.  In many cases, the mere fact that a potential victim is armed at all and willing to brandish his weapon is enough to thwart an attack, even if the gun is never fired.  Thus, having a small-caliber, "underpowered" pistol is infinitely superior to having nothing at all.  That said, the bore size of a .45 caliber pistol is likely to be somewhat more intimidating than the bore of a .22.  The psychological effect may very well be something to consider.

Let's suppose the mere presence of the gun isn't enough to deter an assailant and you have to pull the trigger.  Will your cartridge stop the bad guy?  It depends.

If you shoot and miss, the realization that he's actually being shot at may shock some sense into him, but he'll still be unscathed and fully operational.

If you hit him, what happens next will depend on whether you scored a critical hit.

Despite what we see in movies, a human will often not die from a single gunshot wound, and those that do die, rarely die instantly.  Most gunshot wounds puncture important tissues and cause bleeding, which causes loss of blood pressure, which leads to inadequate oxygen flow to the brain, which ultimately causes death.  This generally doesn't happen right away.  It can take quite a while, actually.  For this reason, trained law enforcement snipers will generally target a criminal's central nervous system (CNS) rather than his vascular system.  While a solid hit to the vascular system can cause death eventually, a critical hit to the CNS will cause instant shutdown.

Unfortunately, the chance of a civilian scoring a critical hit to the CNS of a moving target with a pistol at any distance at all is somewhere between slim and none.  Thus, if you get into a gun fight, you will likely be relying on a different "stoppage" mechanism.