East Precinct Captain Paul McDonagh recently offered this set of data from March crime trends:
- Robberies down 25%
- Strong-arm street robberies continue a downward trend
- Residential burglaries are down
- Business burglaries are down
- Auto theft down after several prolific auto thieves were put in jail
- Car prowls are down 50%
- Aggravated assaults are up
It’s good to hear but would be even better if there were publicly available data to back it up. With numbers we can find to “back it up,” we’ve found that Capitol Hill had 6% fewer emergency 911 call-outs for the Seattle Fire Department so far in 2009. With all of the attention that neighborhood blogging gets when people discuss ‘hyperlocal’ journalism and the future of media, the Web as local data repository gets short shrift. Information about your neighborhood is piling up and, thanks to cool tools and smart humans, more and more of it is being organized into useful formats. Some of that we will call “news.”
While it doesn’t — yet — have crime stats, EveryBlock is one service starting to bridge the gap between information and useful formats. I consult the EveryBlock Capitol Hill page regularly to get a quick update on the various information sources it monitors and collects. I’ve also found a way to collect historical numbers from the system to create trends for a few of EveryBlock’s info services. Here is a look at what I found and what I hope to continue following over time.
The first EB dataset I looked at is the number of 911 fire department call-outs in our area over time. Here is a graph I created from the data. I was interested to see two things. One, did the city’s much-maligned response to January snow result in more emergencies than last year (it appears to have not) and, two, are we seeing more callouts overall this year than in 2008? A hypothesis could be that, if crime is up because of the bad economy, we might be seeing increased emergency medical responses. We’re not. At least, not by this measure. In fact, through the first quarter of the year, emergency responses were 6% lower than 2008.
Another dataset that can be extracted from EveryBlock is the number of building permits issued for things like electrical work in the neighborhood. Below is a graph of the total permits tallied by EB for the first three months of the year in 2008 and 2009. Permit totals are clearly down year over year but March’s total shows some relative improvement. A sign of improvment? We’ll watch this total to see if it’s a useful proxy for economic activity in our area. Overall, permits are down 21% vs. 2008.
One last EB measure is the count of news reports the system has categorized for our area. EB didn’t collect this information in 2008 so no yearly analysis is possible but we’ll keep our eye on this through the year to see how much attention the neighborhood is getting.
Of course, this article might get organized by EB into the Capitol Hill media pile, thus inflating the total with an analysis of those very totals. If that doesn’t rip a hole in the Cap Hill time space continuum, we’ll update this analysis again next month.
the police data does seem significant. too bad we don’t have a gander at the backing data.
6% in 911 calls doesn’t seem quite big enough to be significant in a year over year. but it’s interesting.
building permits data is definitely significant.
i personally wouldn’t really trust comparisons of news reports year-over-year. that’s too subjective and prone to bias. maybe petty crime gets over-reported this year because of the focus on economy (just as example). maybe the data changes dramatically from last year to this year due to the death of the P|I. too many uncontrollable variables.
what’s the equation for interestingness?
interesting as distinct from statistically significant :)