Subscribe to DSC Newsletter

Chicago Homicides 2017, a Follow-Up Look with R

Guest blog post by Steve Miller.

A year ago, I posted an article on the disturbing 57% increase in Chicago homicides for 2016. There's been no shortage of loaded commentary since, including strong statements by the POTUS. A bit more balanced prerspective was provided by fivethiryeight.

There can be no denying the off-the-charts Chicago homicide numbers -- 493 in 2015, 777 in 2016. But while Chicago had far more murders than other U.S. cities, New York and LA included, the murder rate (homicides/100000 population) was exceeded by several other metropolises, including Detroit, New Orleans, St Louis, and Baltimore.

Al Capone

With daily access to Chicago data (after a 7 day lag), I kept up with Chicago crime throughout 2017, looking for rays of hope in the data. Yet while there was a decline of over 100 homicides from 2016 to 2017, I was in no way inclined to claim even a minor victory. After all, the final 2017 number was still 175 more than in 2015 -- so the decline could simply be regression to the mean.

What follows is a look at the 2001 through January 2017 Chicago homicide data, embellished by additional numbers from wikipedia.

The technologies deployed are JupyterLab with an R 3.4 kernel. The scripts are driven primarily through the R data.table and tidyverse packages. Hopefully, readers will see just how powerful these tools are in collaboration. Notable is that neither data.table nor tidyverse is a part of "core" R; each is an addon maintained by the energetic R ecosystem.

The full article, with commented source code (in R) and charts, is available here.

Views: 1682

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Pete Mancini on February 20, 2018 at 11:33am

It would be interesting to see a plot of the blocks where the murders occurred to see if particular neighborhoods that had been quiet suddenly were erupting in violence. It's hard to justify a 56% increase in any data with a really large sample size (in this case population ~2.7M). From what I have read, 5 of the 77 communities contributed 50% of the increase. Austin, Englewood, New City, West Englewood and West Garfield Park. Some claimed it was teenage violence on the rise but the average age of the offender was 26. Also, its hard to attribute it to gang violence, because gang affiliation in offenders dropped from 73% to 67%. There was no surge in availability of illegal guns, according to crime statistics for Chicago. The weather was not exceptionally hot. Poverty wasn't on the rise to explain it, as well. Gun violence did not keep pace with other means of homicide. While it was up, it did not grow at the rate of other forms, such as blades, and blunt force weapons. What isn't in your analysis is a look at current events and policy changes.

Current events during the period: The police shooting Laquan McDonald in 2014 and subsequent release of the video. The Justice Department then began investigating civil rights violations at CPD.

Policy Changes: due to the civil rights investigation, street "stop and frisk" activity by CPD was cut by 80%.

A sentiment analysis of social media posts from those 5 precincts might give insight into the cause of the increased violence. It may also be possible, sentiment was only the match. Perhaps, once the violence hit a certain level, the game theory changed and non-violent means of conflict resolution became undesirable.

Follow Us

Videos

  • Add Videos
  • View All

Resources

© 2018   Data Science Central™   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service