Disclaimer: The discussions made in this post are solely my opinions based on this data project’s conclusion. Additionally, while I have made every attempt to validate the data visualization, I cannot guarantee its accuracy. Please do independent research from other sources before coming to a consensus.
One environment that’s often overlooked when people think of the outdoors is the urban environment. While it is usually what people want to escape from when they go outside, it’s still an environment that many individuals spend their time in. One disruptor of this environment is crime. Sure, crime does not only happen in urban environments—however, crime figures are generally correlated to population size.
Recently, I looked at Charlotte’s crime through public access to CMPD’s crime data. This free resource documents every Charlotte crime and non-crime incident from 2017 onwards. It’s generally updated a couple of weeks before the current date.
I pursued this for two reasons: I was looking for an interesting data project, and this is a resource that was begging to be visualized. Also, this data is genuinely important to analyze and understand as it affects citizen’s day-to-day lives and can help uncover deeper issues of why crime happens.
From this, I ended up building a visualization to show Charlotte’s crime, particularly Charlotte’s violent crime. Additionally, I incorporated US Census Bureau data to help show some details other than “this area has a ton of violent crime”. This can be seen in the map view of Mecklenburg County, which is broken up by census tracts.
What I ended up building shines some light on violent crime in Charlotte. Why violent crime? Well, it’s the one type of crime that directly makes an area hostile and disrupts the peace of its people. Other crimes can do this as well, but violent crimes are the worst of them all. Below you will find the Tableau Public visual I made. Just FYI, the visualization is a bit easier to interact with on Tableau Public’s site.
Thoughts and Conclusions
After exploratory analysis of the data, I noticed a few interesting things about Charlotte’s crime that is violent:
- Areas to the north and west of Charlotte’s city center have the largest volume of violent crime
- Charlotte’s violent crime is generally high overall compared to other major metros.
- Violent crime follows the seasons, mostly in the summer.
- There seems to be a trend where more violent incidents are staying open.
Most of these items on Charlotte crimes that are violent seem to be expected. The one about Charlotte’s larger rate of violent crime compared to other major metro cities is tough but expected. Charlotte does rank higher than average for overall crime, not just violent (source). But Charlotte crime in general and Charlotte violent crime isn’t an emergency that citizens should feel the need to worry about. The majority of the city is safe, and general precautions when going to Charlotte will keep you safe the vast majority of the time.
I do want to draw attention to the open rate in my Tableau visualization. I’m not well-versed enough to know specifics, but this does alarm me. It looks like sometime after 2022, the number of crimes left open., based on dates reported, is consistently increasing. From my understanding of the NIBRS system (learn more about it here), this means that an offender was not found for the crime. This has nothing to do with non-prosecution or that the crime was a false report. Simply put, the person who committed the crime has not been located and/or arrested.
This trend could be due to several reasons. One is that possibly the newer the crime, the less likely it is to be closed. This would make sense, but I would think the trend would show up in a much shorter timeframe. Maybe 3-6 months. Beyond that, I haven’t gotten a clue. Maybe resources are an issue, or maybe it’s a less-critical type of crime to solve. This warrants some further investigation that I may embark on.
Development and Data Sources
With anything data-related, I always want to get into the data I used, the assumptions made, and how I prepped the data. First, the incident data came from the City of Charlotte’s Data Hub. This resource features multiple public data sources relating to the City of Charlotte. I used CMPD’s incident data as referenced in the dashboard.
For census data, I used ACS 2022 5-year data. censusreporter.org was a great resource to help me understand this data. The U.S. Census Bureau’s data repository is a bit clunky to use. Census Reporter helped me to understand the different data sources and tables, how it has broken out, and even if the tables were uploaded for quick. CSV download. Big shoutout to Census Reporter for being an awesome and free resource to use.
As far as data prep, I had to assign each incident a census tract based on the coordinates recorded. GeoPandas in Python as well as the spatial files were necessary here. From there, I could easily join the census data and crime data into one big CSV file.
Tableau Public was easy to use for this purpose. I spent the majority of time figuring out what was going on in the data as well as validating it before I began to build this dashboard. There’s a ton of other information hidden in the CMPD incident data. Combined with the census data, you can begin to look at trends and possibly predict crime. But this is where it gets out of my wheelhouse.
In Conclusion
I genuinely enjoyed this project. Sure, it’s not a light subject, but it’s a real one. So, shining some light on the information for myself and others is enjoyable. If you have any thoughts on the data or notice issues, please let me know! I’m always open to learning about other people’s experiences with data like this. Additionally, I hope to add more data projects here. The census data will become a major piece in some work I may do in the future. So, I’m glad I’ve gotten my hands dirty.
Leave a Reply