Is Chicago really the most dangerous place in America? — Comparative EDA

While researching online, I came across this article making a strong causal claim regarding the most dangerous city in America. We have all heard of the classic stereotype that Chicago can be a violence place, coining it the nickname “Chiraq” due to a rise in gun violence and gang activity, but is it truly as bad as everyone says it is? In order to try and find out, I am going to look at what the data says. I will be using a comprehensive record of over 260k U.S. gun violence incidents between 2013 and 2018 from Kaggle and use EDA processes to investigate whether or not these claims remain valid and consistent.

Link to article: https://www.reckontalk.com/welcome-to-deadliest-city-of-america-chicago-chiraq-usa/

You can see from these two images of the article above that the rise of gun violence due to gang activity and the reported easy access to a variety of firearms throughout the community. Next, we will look into our dataset to see if we can view these trends as well. The first thing we will do is load up our imports.

Let’s take a look at what this dataset has to offer before we continue further.

From our initial look at the dataset we will be using, we can see that it has records of the timestamp, state, county, and the number of people who were either (a) killed or (b) injured due to gun violence. There are also data columns of the victim’s age and gender demographic information, but we will focus on the chronological and geological attributes within this dataset. To begin our EDA process, I decided to use a .groupby() function for the killed and injured reports by year into a stacked bar chart.

Something that you can see from this dataset is that there is a significant drop from 2017 to 2018 in this plot visualization, so my assumption is that it is due to the lack of complete data for the 2018 year, as the last data records only data to (03–31–2018) which tells me that we cannot make accurate assumptions based off of this year due to the lack of information. From here, I used an sns.barplot between the “killed” and “state” then comparing it next to“injured” and “state” barplot to determine if we can see any interesting trends to see the validity that Chicago is the most dangerous city of America.

The plot visualization above shows some interesting trends that I was not expecting. Initially, I expected Illinois to have the highest sum of injured and killed plots, but Chicago believe it or not, ranks as the fourth highest. This brought me to the question, why doesn’t the gun violence by injuries chart resemble the deaths by gun violence chart? What makes California more prone to firearm casualties compared to more frequent locations such as Illinois? These are questions that I would be interested in investigating, but the only conclusion I can currently make from the data is that, Chicago appears to experience more shootings on average than any other city in America, but it is not the most dangerous city in terms of deaths by gun violence. This result is surprising to me, mostly because I don’t fully understand the correlation between death and injury rate occurrences based on frequency. The next step I wanted to do for this EDA process was to visualize the states who have the highest deaths by gun violence by year. I used similar plotting strategies but segregated each sum result to be set only by year instead of a collective cumulative result.

This collection of charts indicate that Chicago is in fact typically among the top three highest states for death by gun violence. It is difficult to claim that Chicago is objectively the most dangerous place in America, as it often varies between California, Texas, Florida, and Illinois. This EDA likely lacks the ability to account for all the other unknown potential variables that could make a city dangerous, but this collection of data for gun violence in America does show indications that could be used to approve or dispute the argument entirely based upon what specifically is considered “dangerous”? The volume of fatalities by firearms throughout states fluctuate too heavily to accurately make a long-standing conclusion.

So, is Chicago really the most dangerous place in America? Well, yes and no. It does rank the highest for the most records of individuals injured by gunfire, but it does not rank the overall highest for highest death rate from firearms. Chicago is a beautiful, vibrant and inspiring location that holds a rich history and deep appreciation for their individual culture, but the title of being “the most dangerous place in America” comes with reason, but may not be entirely correct looking at the situation objectively.