Archive Logo.jpg

May 02, 2006

Murder in NYC

Moving on to other news... how 'bout that NYT? They did some useful analysis, and must have *hated* writing the story to go with it. They analyzed 1,662 murders in New York City from 2003-2005. Summary chart available here. (if registration is required and you don't want to - go to

New York Killers, and Those Killed, by Numbers By JO CRAVEN McGINTY Published: April 28, 2006 The oldest killer was 88; he murdered his wife. The youngest was 9; she stabbed her friend. The women were more than twice as likely as men to murder a current spouse or lover. But once the romance was over, only the men killed their exes. The deadliest day was on July 10, 2004, when eight people died in separate homicides.

Five people eliminated a boss; 10 others murdered co-workers. Males who killed favored firearms, while women and girls chose knives as often as guns. More homicides occurred in Brooklyn than in any other borough. More happened on Saturday. And roughly a third are unsolved.

At the end of each year, the New York Police Department reports the number of killings — there were 540 in 2005. Typically, much is made of how the number has fallen in recent years — to totals not seen since the early 1960's. But beyond summarizing the overarching trends, the police spend little time compiling the individual details.

It's really a very useful story - and the Times did a more thorough analysis of the statistics than the NYPD routinely does. Which I understand to a point - the NYPD wants to determine patterns and develop police responses.

The offender and victim were of the same race in more than three-quarters of the killings. And according to Mr. Farrell, they often had something else in common: More than 90 percent of the killers had criminal records; and of those who wound up killed, more than half had them.

I wonder how many New Yorkers take *this next* message away - given how the news is reported, both print and television (this particular story displaying one of the strengths of print journalism, btw, depth and persistence - you can just study it, not have to keep replaying a video...).

"If the average New Yorker is concerned about being murdered in a random crime, the odds of that happening are really remote," Mr. Farrell said. "If you are living apart from a life of crime, your risk is negligible."

I think the piece is really pretty well written, especially for coverage of this particular issue. There is a *screaming* "however," though....

The offender and victim were of the same race in more than three-quarters of the killings. [emphasis mine] And according to Mr. Farrell, they often had something else in common: More than 90 percent of the killers had criminal records; and of those who wound up killed, more than half had them.

That's the only place race is mentioned in the article. If the article is intended to inform, influence, and shape public policy, that would be okay - if race was a neutral data point. Here, however, it isn't. If you want some bang for your policy investment dollar, you need to target it, rather than develop one-size-fits-all templates. One reason drugs are more effective these days (*and* have more side effects it would seem) is because they are more refined, more targeted - in other words, people are different. Having three drugs available for a medical problem quite possibly accounts for differences between people, hence a drug works for you, but I get all the side effects - that other drug works for me. Same it true of one-size-fits-all public policy.

What am I ranting about?

Roger Clegg pegged it on National Review Online's The Corner:

While whites and Asians are underrepresented among murderers and murder victims in New York City, it does not appear to be the case that Latinos are overrepresented; in fact, their murderer/victim rate is the same as their percentage in the general population (about 27 or 28 percent). The overrepresentation is among African Americans, who are 25 percent of the population, but make up 60 and 61 percent of murder victims and killers, respectively.

The NYT skips that unpleasant little bit of analysis. But if you want to target your policy for effectiveness, in NYC at least, this tells you where to place your effort.

And it further highlights the seeming trend that inner-city black america is feeding on itself. Whether it's racism, poverty, the outgrowth of a flawed welfare policy, exacerbated by the effects of the Drug War providing a Prohibition-Style economic incentive, I don't know. It's *all* inter-related, I'm sure. But I strongly suggest that maybe, just maybe, Bill Cosby is right, and Jesse Jackson and Al Sharpton are not. While there is much that society can do... the real impetus for change has to come from the sub-culture, aided and abetted in context by the larger culture it nestles in. It's a problem we all share - but ignoring the elephant in the living room isn't going to help it.

But we aren't allowed to say that, are we?

And we can't assume that what might work in NYC will work in San Antonio, Kansas City, Reno, or Los Angeles. Hence, the value of a "federalist" approach, coupled with in depth, honest analysis, vice either an "imposed from Washington" or "let's copy the NYC response" and make this generic template fit brute force approach that hasn't worked in the past...

You'd think the Left would embrace this. The brute force template approach assumes assimilation and acculturation, a concept anathema to them (well, except when it means conformance with their template). But, I'm guessing that's not how they want to work it - because it's just so much easier to use the Federal Government and the Courts to make the world look everywhere like you want it - especially if those hick locals don't want to have a Big Eastern City flavor to their midwestern town.

I have to admit - the big stick *is* appropriate at times. It took the Big Stick to knock apart the structure of Jim Crow. But the Big Stick isn't *always* the answer, and a more subtle and nuanced approach might well produce better results.

The whole story can be read here.