Saturday, January 26, 2008

SPD Crime Data: ?

A week ago, the mayor and the police department were crowing about how the Seattle crime rate has reached a 40-year low.

At the top of the list, they pointed out that the murder rate has dropped 20%, which didn't seem to jibe with the number of homicide incidents I heard about last year.

Oh, yeah. In 2006, Kyle Huffs killed six people one night on Capitol Hill. If you were to disregard that one unusual mass killing, the murder rate reported by the SPD from 2006 to 2007 would have been static, at 24 people each year.

But there still feels like there were more murders than that. Did the SPD forget some?

I've been looking at the monthly census tract data for 2006 and 2007, and I don't see some of those who were killed around here last year.

The Seattle P-I reported a murder (and nearby suicide) on December 15, 2007, at a "house in the 8600 block of Beacon Avenue South."

Yet when you enter an address in the mid-8600s in the city's neighborhood crime statistics website, it shows you that there were no homicides in that whole census tract that month:

SouthBeaconHomicide2007

And the P-I also reported that my neighbor was shot and killed in his front yard in the 2100 block of S. Graham Street on September 9, 2007. (That note about how the P-I originally misidentified the neighborhood was put up after I chewed the reporter out for calling it Rainier Valley.)

Yet when you enter that address (I know the exact address, so that's what I used) in the city's neighborhood crime statistics website, it shows you that there were no homicides in that census tract that month:

MidBeaconHomicide2007

I'm not saying there's anything fishy going on. I'm just saying that the city seems to be showing us bad data, and I'd love to see the real information instead. I'll e-mail my neighborhood's SPD liaison Mark Solomon ask him what's up with this.

5 comments:

K and/or J said...

Fascinating. Interested to see what you find out.

Seattle Crime Blogger said...

Interesting point indeed, but we shouldn't jump to conclusions.

For all we know - and this is just pure speculation on my part - deaths could be counted based on where the victim is actually declared dead. In that case, the reason you don't see any murder's in census tract 117 is because the death could have been declared in an ambulance or the ER. Another possible explanation is that the city's neighborhood stats website takes a while to update (which would explain why December murders have yet to appear). The reason I suspect this is that the main neighborhood stats site still reflects 2006 statistics.

Again, pure speculation. But we should remember that a recent surge in crime doesn't mean that crime overall isn't down. I doubt the city would be touting their low crime rates if there was any possible chance that those stats could be wrong.

K and/or J said...

Yeah, seems like this is likely a reporting problem. Main concern is that the numbers are collected/reported in such a way as to intentionally minimize some key measurements. But agree it's unlikely.

JvA said...

Good theories. But in both the P-I stories I quoted, it's reported that the victims were pronounced dead at the scene; they didn't die in the ambulance. And I don't think it's merely a update issue: one of them died back in September, while one of the charts lists a homicide in October.

I believe them when they say the overall homicide rate is down year over year, as we had no mass killing in the city in 2007, as we did in 2006. But their published neighborhood figures do seem off so far.

Mark Solomon wrote me back today with this:

"I'm not sure why those aren't showing up. On our website, they would just show up as an event in a census tract with no identifying information (address or name) associated. You can try our public disclosure section at 684-5481 to get some more info."

JvA said...

Also, if neighborhood homicide rates were documented according to which ERs the victims died in, wouldn't First Hill show a wildly disproportionate murder rate? :)

I'll see if I can get to the bottom of this because now I'm really curious how they come up with these neighborhood stats.