National Crime Rates

Do crime rates decrease as median income and graduation rates increase while poverty level decreases? This study attempts to address this question by examining 5 cities throughout the United States, as well as the national average for the U.S.

We pulled data from the FBI crime database and the Open Data Network for 5 different cities: Tampa, New York City, Los Angeles, Chicago, and Washington D.C. To do this, we utilized a series of API calls using JSON formatting. Once we had our data, we separated it out by city, ran statistical analyses on each of them (to locate any discrepancies in data), and cleaned/standardized the data before combining the various cities into a comparison database along with the national average.

 
 
Violent Crime Rate Comparison (Line).png
Violent Crime Rate Comparison (Bar).png

Above are the year to year violent crime rates of Tampa, Los Angeles, Chicago, Washington D.C., and New York City. You can see that the crime rates dip and increase over the course of 10 years, but there are some very important things to take into consideration:

  1. Data for Chicago was only available from 2014 on due to a restructuring of incident reporting and reclassification of crimes.

  2. Los Angeles crime rates quickly increased in 2013 due to a similar reclassification of violent crimes.

  3. The National Average was significantly lower than the rest of the cities, most likely due to the vast number of rural areas that are included in the national statistic.

 
Crime Rate vs. Median Income (Scatter).png
Crime Rate vs. Graduation Rate (Scatter).png
Crime Rate vs. Poverty Rate (Scatter).png

As we can see above, there doesn’t seem to be any strong correlation between crime rates and graduation/poverty level/median income rates. However, that didn’t stop us from running a few statistical tests just to make sure.


Regressions

Violent Crime Rates

Los Angeles: R = -0.186, p=0.563

New York City: NaN

Chicago: NaN

Tampa: R = -0.897, p = 0.0000757

Washington D.C.: R = -0.449, p = 0.143

U.S. Average: R = -0.871, p = 0.000222


Graduation Rates

Regressions

Los Angeles: R = 0.994, p = 0.00000500

New York City: R = 0.992, p = 0.0000119

Chicago: R = 0.993, p = 0.00000848

Tampa: R = 0.975, p = 0.00019

Washington D.C.: R = 0.994, p = 0.0000059

U.S. Average: R = 1.0, p = 1.9206 x 10-50


Poverty Rates

Los Angeles: R = -0.893, p = 0.00286

New York City: R = -0.958, p = 0.000178

Chicago: R = -0.957, p = 0.000187

Tampa: R = -0.912, p = 0.00158

Washington D.C.: R = -0.95, p = 0.000295

U.S. Average = -0.908, p = 0.00177


Median Income

Los Angeles: R = 0.989, p = 0.166

New York City: R = 0.989, p = 2.47 x 10-5

Chicago: R = 0.971, p = 0.000272

Tampa: R = 0.963, p = 0.000482

Washington D.C.: R = 0.988, p = 3.1 x 10-5

U.S. Average: R = 0.828, p = 0.0215

ancova.png
 

Analysis

As we can see from the above regression statistics and then the comparison of those regressions, there isn’t a strong correlation between graduation rates/median incomes/poverty rates and violent crime. While some of the regression lines were useful for determining things like graduation rate and median income year to year, the same could not be said about violent crime rates. This could be due to several outlying factors, such as the way that violent crimes are categorized per city and if incomes were collected per individual or per household.

Since the cities themselves don’t correlate very well with each other, we can conclude that it would be better to examine an individual city by itself instead of comparing cities against each other, as several societal and economic factors could come into play that would skew the statistics. If we were to examine a city by itself, the rates could be broken down by neighborhood and then a relationship could possibly be seen between graduation rates/median incomes/poverty rates and crime rates by geographic area.

Previous
Previous

Mars Datascrape

Next
Next

Weather Analysis