Lies, Damn Lies, and MPD Statistics
Riverwest and Lower East Side Crime since 2005
Once again, MPD announces positive citywide crime "trends" and long-term decreases without a critical word from journalists or criminologists in the Journal Sentinel's version of the story. (WUWM talked to criminal justice scholar and dean of UWM's Bader School of Social Welfare, Stan Stojkovic.) While I am happy to hear of any decreases, these new "statistics" verge on being meaningless for lack of context and by being reduced to summary form from much more detailed sources that are not communicated to the public.
Since no one lives in all of the city, citywide crime totals have no bearing on anyone's actual lived experience or quality of life whatsoever. Far more specific geographic distributions of the same data are available, so it is possible to see what is going on at the level of police and aldermanic districts, neighborhood strategic planning areas (NSPs), and census tracts. Specific address ranges (one or both sides of a given length of street) or distances in feet from a specific address can also be queried for crime data from January 1, 2005 to the present.
Using this data for NSP 7 (Riverwest and the Lower East Side), here is a "motion chart" showing the relative quantity and change in crime incident types since 2005. (Click the "play" button in the lower left corner to set the animation in motion. Different display options can be selected to alter the way the data is represented.)
While five years of data is not a basis for predicting a long-term trend, there is a general decline visible after general increases in 2006-2007. But after that the frequency of homicides, sex offenses, arson, and burglaries go up, hold steady, or are on pace to increase this year.
With even more spatially targeted data, it is possible to see a doubling of robberies within 2,000 feet of 926 E. North Ave. in 2007 -- and a tripling of their frequency that year from May 1 to September 30 -- although it does not appear as the slightest a blip in the total number of robberies for NSP 7 as a whole.
Clearly location matters a lot, but the statistics MPD parades before the public and the press are not only devoid of meaningful geographic context, the data itself is poor.
MPD reports crime data to the Federal Bureau of Investigation (FBI) and State Office of Justice Assistance (OJA) in the form of detailed, comprehensive Incident Based Reports (IBR). The Wisconsin IBR System (WIBRS) is patterned on the National IBR System (NIBRS). (You can access NIBR data and much else at the University of Michigan's Archive of Criminal Justice Data.)
Incident-based reporting was developed in the late 1980s to replace Summary Based Reporting (SBR), which has been in use since 1929. IBR records a great deal of information about crime incidents (in which more than one type of crime may be committed), while SBR simply uses a single classification for each incident. The crime statistics MPD routinely presents to the media and public are based on SBR data, mainly for Part I offenses in the Uniform Crime Reporting (UCR) classification hierarchy. The main benefit of those UCR statistics is they allow simple comparisons and calculations involving different police jurisdictions. The SBR-based UCR data is less helpful for assessing a single city and police force since in Milwaukee's case it is reductively derived from MPD's IBR records by the State OJA. As MPD's public reports always note, the two systems (IBR and SBR) are totally different; the conversion from one to the other is really apples to oranges. Aside from losing much other data in the IBR, if a single crime incident involves a homicide as well as one or more other violent crimes (rape, robbery, assault, burglary arson, larceny, motor vehicle theft or arson), the SBR classifies it as a single incident of homicide.
What kind of crime data would you find most useful? I see no reason why we can't get better information from a police department funded so well the Common Council pays for its union liaisons and their golf outings.
Crime Graphs for Riverwest and the Lower East Side Combined
(Neighborhood Strategic Planning Area #7)
Does this look like a long-term decrease?
Here's a much more useful representation of block-level crime (for 2008) used for the AreYouSafe iPhone app. Zoom in to see an area up close:
How the crime "heat map" works: The city is mapped out with two-block radius grids. Each grid gets a score based on the severity and quantity of crime incidents in it. Each crime category is weighted by severity. The way homicides impact a grid's score is as 40*(number of occurrences)^3 so that one homicide won't affect the score as much as multiple homicides in the same area will. Then there is an adjustment for population, since the more densely populated areas have more crime. The population of the closest census tract (people per square mile) coinciding with each grid is multiplied with the grid's crime score. A curve is applied to the resulting figure to place the score into one of 10 safety levels, from most to least safe.
Trackback(0)
TrackBack URI for this entryComments (4)
Subscribe to this comment's feedShow/hide comments
...
-
- +0
-
-
@dave
what might be useful is finding the geo-center of riverwest (or other neighborhood) and pulling the data for all crimes, by type, within that radius that encompasses the area you are targeting.
-
- +0
-
-
North Ave. chart
-
- +0
-
-
Write comment
Show/hide comment form
