Got this response directly to my email re: Crime Stats.  Thought critique 
important to share.

--Jennifer Armstrong
Payne/Phalen

First, their population for St. Paul is incredibly low.  That, of course, 
artificially increases the rates.  The history is that, for several decades if 
not longer, the U.S. Census Bureau's annual population estimate for St. Paul has 
shown unrealistic declines.  Then after the decennial census they find out 
that, Surprise! the city was actually increasing in population!, and eventually 
adjust the count to fit the census results.  Unfortunately, the census counts 
take a 1-3 years to be released, so the 2001 estimate was based on anticipated 
population declines since 1990.  However, the crime data they're using was 
released in 2002 and the site was last modified in 2003, so there is no valid 
excuse for using that low population count.  The 2000 count was over 286,000; 
they report 264,000. 

Second, the national average is a pretty meaningless number to compare any 
location to.  St. Paul should be compared to just inner cities, or just urban 
areas. 

Third, crime fluctuates wildly from year to year, and it's pretty meaningless 
to look at a single year's data.  A few events or a crackdown on a certain 
type of crime or a certain neighborhood can make the rates skyrocket.  Any good 
crime analysis will look at many years, or a rolling average. 

Fourth, a city of neighborhoods like St. Paul will have greatly varying crime 
rates in the different neighborhoods.  Many of the neighborhoods are similar 
in size and in safety to the suburbs, but crime elsewhere makes the whole city 
look bad. 

Fifth, crime reporting methodologies can vary depending on the jurisdiction.  
It's a few years since I've looked at this, but when I last checked, some 
areas will report every time the police are dispatched while other areas will 
only report those that were made into an official report.  Anyone who compiles 
these disparate types of reports should either note the distinction or only 
include one reporting methodology.  Unfortunately, few organizations do this. 

While I don't know if all of the above issues influenced the specific data 
shown, I would definitely be aware of those issues when comparing crime data.  
Unfortunately there's no information about these issues in the web site's 
footnotes. 
_____________________________________________
NEW ADDRESS FOR LIST:     [EMAIL PROTECTED]
 
To subscribe, modify subscription, or get your password - visit:
http://www.mnforum.org/mailman/listinfo/stpaul

Archive Address:
   http://www.mnforum.org/mailman/private/stpaul/

Reply via email to