Do guns prevent crime?

In this article, I try to answer one of the most debated topic about guns and crime. Do more guns lower or increase crime?

Obviously there are many who had done this study before. However, due to mixed results and doubtful politically-motivated conclusion, I need to do it myself. That is also true for you or any readers, thus I’ll write in such a way that it is easy for anyone to try to duplicate my results. Plus I’ll throw in my reasoning for data choice and will also try to be as neutral as possible.


 

Comparing U.S. by state

I chose to compare states within the U.S. rather than comparing the U.S. to other countries. That’s because there are too many factors at play when you compare too different groups of people. Culture, income level, conflicts within the country and security level are some of the things that could play bigger roles than gun ownership.

Map_of_USA_with_state_names
Licensed under the Creative Commons Attribution-Share Alike 3.0 Unported; Author : w:User:Wapcaplet:

For example, you wouldn’t want to compare the U.S. with Syria due to conflicts within that nation. Some people suggest European countries or Canada to compare instead. My reasoning is that European countries tend to have smaller areas of land and should be easier to manage in security aspect. While Canada also has large areas of land, it tends to be less populated. Here’s a comparison of population density per square miles. (https://en.wikipedia.org/wiki/List_of_countries_and_dependencies_by_population_density)

  • USA (2017) : 86
  • UK (2016) : 702
  • Canada (2018) : 10
  • Germany (2017) : 601

Not to mention cultural and economic differences. Comparing different countries would be interesting, but it will not answer my question.

Note that states in the U.S. are also vastly different. New Jersey is going to be much more densely populated than Ohio. But at least, states within the U.S. share more of similar cultures, government, income level than comparing to other countries.

 

Data for this Experiment

Here’s my list of data and sources for this experiment.

U.S. percentage of gun ownership by state

I get this data from Wikipedia : Firearm death rates in the United States which got its data from U.S. Census Bureau and FBI reports. This data set provides U.S. percentage of gun ownership by state in 2013.

Percentage ownership is used to better compare states with different population number.

Note that these numbers are approximation due to unregistered guns and illegal guns. That said, the number is likely the best estimate with little to none bias.

 

Violent Crime Rate per 100,000 Inhabitants

Here I use FBI’s violent crime data and definition. Violent crimes are defined as offenses which involve force or threat of force. It is composed of 4 offenses: murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault.

murder crime scene with yellow tape at night

The data describes how many offenses occurred per 100,000 inhabitants in 2016. Note that my gun ownership data is from 2013. There will be some discrepancy on this.

Since the data comes from FBI database, there should be little to none bias. If there’s any bias, the number provided is going to be lower than reality, due to unreported cases or window dressing. (No evidence of such case that I know of.)

 

Property Crime Rate per 100,000 Inhabitants

Here I use FBI’s property crime data and definition. Property crime includes the offenses of burglary, larceny-theft, motor vehicle theft, and arson. The definition of the theft-type offenses is the taking of money or property, but there is no force or threat of force against the victims.

The data describes how many offenses occurred per 100,000 inhabitants in 2016. Note that my gun ownership data is from 2013. There will be some discrepancy on this.

Since the data comes from FBI database, there should be little to none bias. If there’s any bias, the number provided is going to be lower than reality, due to unreported cases or window dressing. (No evidence of such case that I know of.)

 

Number of Key Gun Laws in each State

This is used to answer a different question, “Do stricter gun laws lead to less crime?”. I used data from Everytown Research to determine how many key gun laws each state had in 2016.

Note that counting number of key gun laws do not fully take into account of each laws strictness or effectiveness. An alternative is to use strictness of gun control rating by state. But that would add subjectivity into the equation.

 

Household Income by State

Household Income

In my opinion, poverty should affect crime in a big way. Thus I’ll measure how big guns affect crime compared to a sure fire factor like poverty. 2016 Household Income is collected from the Census Bureau and is used to measure poverty.

Household income is defined as combined incomes of all people sharing a particular household or place of residence. It includes every form of income, e.g., salaries and wages, retirement income, near cash government transfers like food stamps, and investment gains. Put simply, it’s how much a household can spend.

 

Data Table

This is my data table used to calculate.

State Reported violent crime rate per 100,000 (2016) Reported property crime rate per 100,000 (2016) % Gun ownership (2013) Number of Key Gun Laws in each State (2016) Household Income (2016)
Alabama 532.3 2947.8 48.90% 19 46,257
Alaska 804.2 3353 61.70% 7 76,440
Arizona 470.1 2978.4 32.30% 15 53,558
Arkansas 550.9 3268.6 57.90% 23 44,334
California 445.3 2553 20.10% 49 67,739
Colorado 342.6 2740.7 34.30% 26 65,685
Connecticut 227.1 1808 16.60% 44 73,433
Delaware 508.8 2766 5.20% 34 61,757
Florida 430.3 2686.8 32.50% 19 50,860
Georgia 397.6 3004.5 31.60% 13 53,559
Hawaii 309.2 2992.7 45.10% 59 74,511
Idaho 230.3 1744.2 56.90% 6 51,807
Illinois 436.3 2049 26.20% 57 60,960
Indiana 404.7 2589.4 33.80% 21 52,314
Iowa 290.6 2086 33.80% 21 56,247
Kansas 380.4 2695.5 32.20% 14 54,935
Kentucky 232.3 2189.7 42.40% 17 46,659
Louisiana 566.1 3297.7 44.50% 23 45,146
Maine 123.8 1645.7 22.60% 20 53,079
Maryland 472 2284.5 20.70% 37 78,945
Massachusetts 376.9 1561.1 22.60% 55 75,297
Michigan 459 1909.9 28.80% 26 52,492
Minnesota 242.6 2133.3 36.70% 42 65,599
Mississippi 280.5 2768.1 42.80% 5 41,754
Missouri 519.4 2799.1 27.10% 20 51,746
Montana 368.3 2683.5 52.30% 10 50,027
Nebraska 291 2263.3 19.80% 20 56,927
Nevada 678.1 2586.6 37.50% 37 55,180
New Hampshire 197.6 1512.9 14.40% 9 70,936
New Jersey 245 1544.6 11.30% 35 76,126
New Mexico 702.5 3937.1 49.90% 15 46,748
New York 376.2 1545.6 10.30% 53 62,909
North Carolina 372.2 2737.5 28.70% 24 50,584
North Dakota 251.1 2295.9 47.90% 19 60,656
Ohio 300.3 2577.5 19.60% 25 52,334
Oklahoma 449.8 2982.9 31.20% 24 49,176
Oregon 264.6 2964.4 26.60% 33 57,532
Pennsylvania 316.4 1742.7 27.10% 34 56,907
Rhode Island 238.9 1898.7 5.80% 31 60,596
South Carolina 501.8 3243.8 44.40% 24 49,501
South Dakota 418.4 1980.6 35.00% 12 54,467
Tennessee 632.9 2854.1 39.40% 36 48,547
Texas 434.4 2759.8 35.70% 25 56,565
Utah 242.8 2951.5 31.90% 28 65,977
Vermont 158.3 1697.4 28.80% 11 57,677
Virginia 217.6 1859.4 29.30% 28 68,114
Washington 302.2 3494.1 27.70% 34 67,106
West Virginia 358.1 2047.2 54.20% 26 43,385
Wisconsin 305.9 1933.3 34.70% 30 56,811
Wyoming 244.2 1957.3 53.80% 5 59,882

 

List of Relationships to Check

Here’s the relation pair of what I’d like to find out :

  • %Gun Ownership ↔ Violent Crime Rate
  • %Gun Ownership ↔ Property Crime Rate
  • Number of Key Gun Laws ↔ Violent Crime Rate
  • Number of Key Gun Laws ↔ Property Crime Rate
  • Household Income ↔ Violent Crime Rate
  • Household Income ↔ Property Crime Rate

 

Method

Pearson Correlation

Pearson Correlation is often used to determine whether a linear statistical relationship exists between 2 variables. In Microsoft Excel or Google Sheets, you can use this function by typing “=correl(Range1,Range2)”

You’ll get a value between -1 to 1. Positive value means that factor A and factor B tend to increase or decrease together. While negative value means factor A and B tend to go in opposite directions, A increase while B decrease or vice versa. As a rule of thumb, you interpret it like this :

  • -1 to -0.5 or 0.5 to 1 = Strongly correlate
  • -0.5 to -0.3 or 0.3 to 0.5 = Moderately correlate
  • -0.3 to -0.1 or 0.1 to 0.3 = Weakly correlate
  • -0.1 to 0.1 = Little to none correlation

Then you squared the correlation result to get how much % of factor 2 is explained by factor 1 (or vice versa).

  • Suppose correlation = 0.5 ; 0.5 squared is 0.25 which means factor 2 (Violent Crime Rate) tends to go along with factor 1 (Gun Ownership) 25% of the time.
  • Suppose correlation = -0.8 ; -0.8 squared is 0.64 which means factor 2 (Violent Crime Rate) tends to go in the opposite direction with factor 1 (Gun Ownership) 64% the time.

To better analyze the data, one would normally use scatter plot in conjunction with correlation to help determine whether the data has linear relationship. If you plotted them, insert the trendline, and can still see other patterns with your eyes, Pearson Correlation might not be good enough. This is possible as the Pearson Correlation do not detect Exponential, Logarithmic, Polynomial or other relationship.

Note that correlation can’t determine causal relationship. For example, a strong positive correlation between guns and violent crime could mean either of these two :

  • More guns lead to more crime.
  • There are more crime so people buy more guns.

 

Results

%Gun Ownership ↔ Violent Crime Rate

Pearson Correlation = 0.3279

Pearson Correlation Squared = 0.1075

Scatter plot does suggest some other factors at play. But the trend is recognizable in my opinion.

There is a moderate positive relationship between gun ownership and violent crime rate. But as previously noted, causal relationship cannot be checked by this method. More guns can lead to more crime, or more crime can lead to more guns.

We can conclude something though. More guns do not lead to less crime according to this data set. If more guns lead to less crime, we should see negative correlation.

 

%Gun Ownership ↔ Property Crime Rate

Pearson Correlation = 0.4205

Pearson Correlation Squared = 0.1768

Again, there is a moderate relationship between gun ownership and property crime rate. The relationship is stronger than violent crime rate and gun ownership.

If you live in a tough neighbourhood and need some protection, check out our best red dot sight, best red dot for AR, best red dot for AK.

 

Number of Key Gun Laws ↔ Violent Crime Rate

Pearson Correlation = 0.010336653

Pearson Correlation Squared = 0.0001

Apparently, there is little to no relation between number of key gun laws and violent crime rate. Maybe these laws aren’t so effective after all.

 

Number of Key Gun Laws ↔ Property Crime Rate

Pearson Correlation = -0.1683

Pearson Correlation Squared = 0.0283

There is a slight negative correlation suggesting more gun laws does reduce property crime. Even though not by much.

 

Household Income ↔ Violent Crime Rate

Pearson Correlation = -0.2004

Pearson Correlation Squared = 0.0402

There is a weak negative correlation suggesting that more well-to-do people leads to less violent crime. Another interpretation is quite possible as less violent crime leads to more wealth.

Note that the correlation between violent crime ↔ household income is less than violent crime ↔ gun ownership.

 

Household Income ↔ Property Crime Rate

Pearson Correlation = -0.3118

Pearson Correlation Squared = 0.0972

The result shows moderate negative correlation between household income and property crime rate. This relation is stronger than Household Income ↔ Violent Crime Rate but weaker than violent crime ↔ gun ownership.

 

Conclusion

From the data, it seems that more guns do not reduce crime. Even though some might want to conclude that more guns causes more crime, one must remember the fact that correlation offers no causal relation check. The flipside, “There are more crime so people buy more guns.”, can be equally true.

Gun laws do not seem to affect violent crime, but they do reduce property crime. Furthermore, people with more money tend to engage in less crime.

If you have any suggestions to improve my analysis, be it gang crime data, historical data set or a better method to do analyze all this, I’ll be glad to hear.

Back to home

Be the first to comment

Leave a Reply

Your email address will not be published.


*



 

About us

Logo

Hi there. We are a group of shooting/hunting enthusiasts. Our goal is to make shooting/hunting popular again by making things easier for beginners.

 

If you have any questions, comments and/or suggestions, please contact us here.