Pretty Maps, But Not Useful Ones


Note: Will be traveling this weekend, so there will be no Saturday post. I'll try to get back to normal next week, and I'll have a missing person-related post ready for Monday.

Crime author and blogger Stacy Horn recently linked an effort by the NY Times to provide a spatial representation of homicides in the five boroughs of New York City between 2003 and 2009. Representatives from The Times allow a viewer to plot the crimes by several categories including temporally, sex/race of the victim and perpetraitor, and if a weapon was used.

The reader can also zoom into street level to see approximate locations of homicides.

Under the well constructed GIS work, the following text appears:

Do you see a pattern that should be explored further? Have suggestions for a story? Comments about this database? Please email us.
Well they asked so here would be my comment:

I love maps as much as the next guy (sorry I am stereotyping), but this is nothing more than a lot of time-consuming work with little potential for substantive analysis. In its current state, you lump every kind of homicide together; when there is plenty of variation within that category.

For instance, there is a distinct difference between the meaning of a domestic related homicide (to a citizen) involving a husband and wife and one homeless man killing another over a liquor bottle. Yet each of these crimes would represent the same two blue dots on the displayed map.

If you really want to crunch some numbers and make the spatial data useful, start categorizing the homicides by typology (some may be unknown and that is ok). I want to know how many domestic violence murders occur in the Bronx. Where are the hot-spots for drug related killings? Should residents of lower Manhattan be concerned with homicides that began as armed robberies?

The FBI’s Uniform Crime Reports use classifications like homicide to track police reported crime in the United States—it does not mean that you as newspaper reporters are stuck with that standard.


One of the professors (and a former officer) that I hope to work with in the future, told me a humorous story about assisting a large city’s police department with evaluating their juvenile crime problems. After spatially analyzing the agency’s use of curfew enforcement to sweep juveniles off the streets, he found that the highest juvenile crime rates at night were located in the areas where the teens were being returned home (after being taken into custody by police) as opposed to their arrest location.

The professor had a really fun time informing the chief that he may need to rethink his curfew enforcement strategy—making it sound better than “hey, chief, you all are better off just leaving the delinquents on the street corners where they hang out.”

In addition to providing an example of a good spatial analysis, I think the story unfortunately also illustrates the troubled home-life of a significant number of urban young people.

With the Times crime maps, I concluded with this: nice try, but there is plenty of room for improvement.

5 comments:

Raindog said...

Great Map!. Thank you for posting it.

J. J. in Phila said...

I think that you are correct. One other think that should be looked at is transportation and foot traffic. High traffic areas might have more crime, because more people pass through a given area.

Katherine Mercurio Gotthardt said...

"he highest juvenile crime rates at night were located in the areas where the teens were being returned home"

That's a pretty telling and sad statement, isn't it?

mappchik said...

I like the map, but find myself asking what the point was for publishing it. Sure, it's interesting to see the homicide data mapped out, from a "hey, look at this" standpoint.

It almost looks like the entire piece was published as a fishing expedition by the writer. Each person who writes in with ways to filter the data to look at specifics will be helping out with the research for future articles.

That being said... I like your comments, and think they'd spark several very good pieces which could come from the spatial data.

mappchik said...

Oops - not wondering your point in posting map, but the NY Times writer.