Do Relaxed Gun Laws Lead to More Shootings?
Gun deaths have become so common in the United States that we are often numb to them whenever the next one comes along. With each of these comes a new cycle of public outrage, heartbreak, and political gridlock. Some argue that increased government policy will lead to a decrease in shootings. Others argue that the perpetrators of these crimes are so evil they will obtain weapons regardless of government policy. So the question becomes: without getting too bogged down by political biases, how does the implementation of gun laws affect the prevalence of mass shootings? I’ve acquired three data sets to help us answer this question.
The Data
Two of the three data sets used for this project were found on Kaggle.com. The last was published by Gifford’s Law Center. Links will be provided to each.
- gun_violence_df: A comprehensive record of about 260k mass shootings that occurred in the US from late 2013 through early 2018.
- population_df: A record of each state’s 2018 population. Does not include District of Columbia. Used to generate per-capita statistics.
- laws_df: A letter grade rating of the strength of each state’s gun laws. Published by Gifford’s Law Center, a group aimed at preventing gun violence.
To create the visualizations used later in this piece, I created a master data frame which included features from all three of the above data sets. To make gun_violence_df merge-able with the other two, I used group-by aggregation to boil down the total number of incidents and total people harmed per state in the data set. I then joined the three together using the ‘State’ column.
Additionally, the data for all incidents pre-2013 is incomplete in this data set. For that reason, I’ve filtered the set down to all incidents from January 1st, 2014 and onward.
Common Gun Debate Biases
Many studies have shown that political bias harms our ability to use logic to arrive at objective fact. A study conducted shortly after the Sandy Hook shooting demonstrated that most intelligent adults were able to use logic to correctly answer a simple business question, but those same adults were not able to correctly answer a similarly-structured question about politics. The researchers concluded that, “political bias had erased the advantages of stronger reasoning skills.”
There are many other biases present within this debate as well, such as the Endowment Effect. All of us hold this bias, as we are more likely to place a higher value on things simply because we own them. As such, Americans that currently own guns are more likely to oppose policy aimed at taking them away, as losing this item may make them feel more uncertain about the future.
Additionally, both Liberals and Conservatives are guilty of the Representativeness Heuristic. We as humans tend to make generalizations about large amounts of information. Additionally, it may lead us to believe that the most extreme examples of a certain ideology are representative of the entire group. For example, Liberals may jump to identify all American gun-owners as violent people, even though a very small portion of that demographic has committed a violent crime.
Analysis
To identify whether stronger gun laws had an impact on decreasing shootings, I combined my three data sets together. Gun laws are generally left up to state governments, so I’ve taken our massive list of individual shooting incidents and condensed it down to per-state averages.
Additionally, I used some data within the existing columns to create new metrics. These metrics, such as people harmed per incident, number of incidents per capita, and gun death rate per capita allowed for more in-depth analysis. Creating per capita statistics also allows us to compare states of varying population size in a fair manner.
To narrow the focus of my analysis a bit, I chose to focus on mass shootings only. There is no widely accepted definition of what a mass shooting truly is, so I was free to create my own to some extent. For the purposes of this project, I chose to define a mass shooting as an incident where six or more people are injured/killed in total. Any incidents with five or less people harmed were left out of the following analysis.
To see how gun laws affect the prevalence of mass shootings, I took the top-ten worst states in a number of gun violence related categories, and saw how they scored in Gifford Law Center’s grades. The results were not surprising, to say the least.
In the visualization above, we have the top ten states in terms of the number of mass shootings per 100,000 people. Not surprisingly, these ten states have some of the most relaxed gun laws in the union, per Gifford. Five out of the ten listed scored an F, and seven out of ten scored as a D- or worse. Early on, we are starting to see a correlation between lax gun laws and increase in shootings. Illinois is the main outlier here, as they still have a high instance of mass shootings despite relatively strong gun laws.
Additionally, I studied which states tend to have the most people harmed per mass shooting. As was the case with the last visualization, the top ten in this category all tend to have very relaxed gun laws. Once again, we must note that there are outliers here, as California and Connecticut tend to have more deadly shootings despite more strict gun laws. A more detailed analysis of the specific differences in gun laws might reveal the reasoning for this, but unfortunately our data does not have that capability.
Conclusions
This gun violence analysis confirms that there is a correlation between gun laws and the prevalence of mass shootings. States that score poorly in Gifford’s Gun Law ratings tend to have more shootings, and those shootings tend to be more deadly. However, we must remember that correlation does not necessarily imply causation, and that gun laws are not the only factor that influences the prevalence of shootings in America.
To expand upon this research, I’d like to find some data regarding gun registration per state, to see if that also has a correlative effect.
All notebooks and data used in this project can be found here: https://github.com/zjlyons/Gun-Violence
