December 29, 2002
Late And Vague, But Better Than Nothing

This topic is almost two weeks old, and I can't find the crucial reference, but I think I still have something worth saying:

On December 16th, 'Jane Galt' had an interesting post on Asymmetrical Information about "a study arguing that the murder rate has fallen because of better access to medical care. It's one of those unobvious intuitions that make you smack yourself in the head and go 'Duh!'".

Some years back (maybe 5 or 10), the Washington Post had an article comparing murder rates in Washington and Baltimore. I have not been able to find it or a reference to it on the web, but it was an eye-opener. The most striking thing about the article was a graph comparing homicide rates in Washington, D.C. and Baltimore. It looked rather like this:

Over something like 30 years, homicide rates in both cities followed an upward path with a slight but steady acceleration, with one huge exception: along the way, the homicide rate for Baltimore once dropped in half from one year to the next (it may have been 1977). As I recall the story, that was the year the city installed a comprehensive system of on-call helicopters, designated trauma centers staffed with specialists, clearly-defined triage procedures, and up-to-date computerized communications. When the article was written, Washington had still not done so. (In fact, Washington's ambulance service is notoriously incompetent, and there has been at least one case where a heart-attack victim had to help the ambulance driver find the hospital.) The result was that when the story was written Washington residents were roughly twice as likely to be murdered as Baltimoreans, not because the city had twice as many murderous thugs, or because its thugs were twice as brutal or twice as competent, but because people who were shot or stabbed or beaten in Washington were far more likely to die of their wounds than people who were (relatively) fortunate enough to be shot or stabbed or beaten in Baltimore.

In other words, it's not just the increase in the competence of emergency-room doctors and the overall quality of their equipment: that process must always be fairly gradual. It's the ability to get victims of attempted homicide to the Emergency Room quickly, to keep them alive on the way, and to have the doctors, nurses, and equipment ready to deal with them as soon as they arrive. That takes a well-organized system.

The intersection of criminology and medicine is not my primary field of interest, but the article would be well worth tracking down if it is yours. Too bad I have no idea who wrote it, or in what year.

Posted by Dr. Weevil at December 29, 2002 06:28 PM

The likelihood that the stability of US murder rates over the past few decades is due to improvements in emergency care was discussed in brief in the New York Times Sunday Magazine on Dec 15.
The frightening corollary, however, is that there has been a huge increase in murderous assaults over the same period. The stats for aggravated assault have indeed shown such an increase, but most of us have been lulled by the apparently more significant murder stats (reporting of which is also necessarily more consistent and reliable).

There is also an important philosophical issue here, which I'd love to see legal scholars tackle online (Volokhs, perhaps?): should we continue to punish according to the results of a criminal action, or according to its intent? Certainly we allow consideration of intent as a mitigating factor, and in some narrow instances as an aggravating factor, as well. But overall, our system will still punish a murderous attack very differently depending on the outcome, which may have nothing whatsoever to do with the attack itself.

Posted by: Cronaca on December 30, 2002 12:31 PM

It's not at all clear whether the apparent increase in serious assaults (whatever they are called) is due to an actual increase in violence, or to a much lower threshold for getting an altercation classified as a felony rather than a misdemeanor, or to be reported at all. That is, 30 years ago cops would break up a fight, but if no one was hurt much and no one wanted to file a complaint they probably wouldn't even arrest anyone. (Unless one of the participants was recognized as belonging to some group such as gang members, habitual criminals, or hippies that the cops would run in at every excuse.) But now and then they'd get a little trouble over that, because the next night the same guys might bring real weapons and renew the fight, with someone winding up in the ER. So now, they'll probably not only cover their rears by arresting everyone, but also report it as a serious (felony) assault just to get more leverage for getting a plea bargain to a misdemeanor. (Arrests probably _increase_ the chances the fighters will try to kill each other as soon as they make bail, but the cops are in the clear legally.)

Now if you can find stats on assaults in which someone was admitted to the hospital for at least overnight, then you might have semi-meaningful data. There may still be a reporting differential in that guys who will claim that injuries that landed them in traction are due to "running into a door" are getting rarer, but it's not likely to be a big part of the numbers.

Posted by: markm on December 30, 2002 07:56 PM

Excellent points! I just saw a stat that the number of murders in Manhattan has dropped to a level not seen since the nineteenth century! I wonder if the same could be said of the number of violent assaults that in the 1890's would have resulted in death but do not now. Also, it should be remembered that murders in the wrong part of town - read black, Hispanic or Chinese, even Italian or Jewish at certain times - were not usually reported to the police and, even if they were, were not considered notable enough to make the papers or the social scientists' statistics.

Posted by: Robert Speirs on December 31, 2002 02:13 PM