GEO 878

Crimes in Boston

Motivation

The United Nations defined a total of 17 sustainable development goals as a guideline to achieve a more sustainable future with the aim to achieve all of them by the year 2030. The idea is to tackle common problems ocurring all around the world such as poverty, inequality as well as climate change or justice (for further information see UN Sustainable Development Goals).
Goal 16 aims to reduce crimes and injustice as these are holding back the development of a society. In order to monitor the current crime situation as well as the recent progress in fighting crimes, the situation in the city of Boston MA was examined.

Compared to the average crime rate across all communities in the United States of America, Boston has one of the highest numbers of crimes per 1000 residents. Nevertheless, the city reveals a good performance when its crime rate is compared to other US-cities of similar size (NeighborhoodScout, 2020). Over the last years, crime numbers have mainly and preliminary decreased (AreaVibes, 2020; MacroTrends).

With the help of this website, the analysis of the occurences of crimes over time as well as the spatial distribution within the city becomes easier.

Project

To be more efficient in the fight against crime and crime prevention, a detailed analysis of the crime situation in recent years is necessary. The aim of this project is to tackle the follwing questions:

Data

In order to make the crime categories clearer for visualization purposes, we have recategorized them into 7 crime type groups. Data from June 2015 to September 2019 are included in the analysis. We grouped the crimes into new groups depending on the attribute “OFFENSE_CODE_GROUP” which had been assigned by the city of Boston:

All crimes which were not labeled in the original dataset or were not included in our recategorization were not included. We chose to focus on crimes that we consider to be crimes that disturb the city’s reputation and often involve. On the one hand, we are aware that for example homicide and manslaughter are legally different crimes (for further information see Murphy law office), but because of their low numbers we decided to group them together in order to not miss out of these important crimes. On the other hand, because Burglary, Robbery and Larceny are represented in high numbers we did not group these crimes together (for further information see criminal defense lawyer).

The other data set that was used for this analysis contained the results of an American Community Survey for the years 2013-2017. It contains data about the distribution of various parameters, such as age, nativity, per capita income or poverty rates. We only included the demographic data of 2017 for our analysis. .

Maps and plots

This map is intended to provide detailed information of the crime density in Boston, especially over time. The desired crime type(s) can be selected, and the map displays the density of the crimes over the defined period of time. The density is displayed regardless the neighborhood boundaries.

The density is calculated by using the kernel density estimation (KDE). By varying the kernel size it is possible to detect small scale hotspots (small kernel size) as well as large scale patterns (large kernel size) of the selected data. While areas with a higher crime density are displayed in dark blue, regions with a lower density are represented in light blue.

By hovering over the different neighborhoods, the number of crimes selected by the settings as well as their ratios per population and per area are displayed.

The density map reveals that from July 2015 to September 2019 there have always been three distinctive crime hotspots in the city of Boston. One is located in “Downtown”, which is also one of the most dangerous neighborhoods according to the choropleth map above. Another hotspot lies on the southern border of “Back Bay”. The last area of high crime density lies exactly on the borders between “South Boston”, “Roxbury” and “South End”. Other areas with local maximum of crime densities also don’t change much over time and remain at the same location. The mentioned areas always represent the highest density relative to other areas. In order to detect how absolute numbers of crimes have evolved over time the next application can be used.

By selecting different types of crimes it can also be seen that their contribution to the density map varies vigorously over time and space. For example on the one hand, Burglary is relatively absent in Downtown at certain times and stronger represented in other neighborhoods, such as Dorchester or Allston. On the other hand, larceny does only show hotspots in Downtown and Back Bay unlike the density of all crimes, which also reveals a high density between South Boston, Roxbury and South End. A good example of a crime that varies spatially over time is Murder which reveals hotspots at different locations at different times.

The map allows you to explore the crime situation in Boston’s neighborhoods. On the right side, a choropleth map is shown which indicates the number of crimes in relation to the population. The legend can be expanded in the top right corner of the map.

The map is interactive:
You can either select neighborhoods using the drop down menu in the upper right corner or activate the cursor symbol in the upper left corner of the map to be able to click directly on individual neighborhoods on the map.

The information on the left side of the dashboard changes depending on what is selected.

The choropleth crime map provides a detailed overview of reported crimes in Boston from June 2015 to September 2019. The colors indicate the number of crimes in relation to the population. The darker blue the neighborhoods are colorized, the more crimes per resident people happened in this neighborhood.

The safest neighborhoods to live in Boston are Beacon Hill, West Roxbury and Brighton. On the other hand the most dangerous neighborhood in respective of crimes is Downtown, which has a crime ratio per population of 0.7. Moreover, Back Bay and Roxbury are unsafe neighborhoods. The demographic indicators in the dashboard show that Back Bay is a neighborhood with a very high per capita income and high level of education. Interestingly, almost 77 % of crimes that happen there are larcenies. On the other hand, Roxbury has a low per capita income and a low education level. A big part of the crimes that happened there are violations and assaults.

By comparing the crime situation and demographic information of Boston’s neighborhood, some interesting knowledge about the cause of crime can be gained. In some neighborhoods, a negative correlation between the level of education/per capita income and crime rate can be observed. A closer analysis of the type of crimes shows that especially violation and assault crimes happen there often. On the other hand, there are also neighborhoods, where we have a positive correlation of the level of education/per capita income and the crime rate. Especially, larcenies happen there often. In other words, the situation in all neighborhoods must be individually analyzed. However, it is important to include demographic information in such an analysis and differentiate between the type of crimes. Additional knowledge can be gained on how to proceed specifically in crime prevention.

By selecting a neighborhood or Boston (all neighborhoods) and the types of crime it is possible to detect the number of crimes as well as the trendline for each type of crime in the plot. Another feature that can be selected is the granularity which allows to display the number of crimes on three different levels (Week, Month and Year). The granularity also defines what time span is being displayed in the plot, since only data from complete years and months are displayed. Additionally, the level which the data is displayed at, gives an idea of the variability of the data and the progress

Over all neighborhoods especially the incidents that are reported very often (larceny or violation) show a decrease in numbers over time and different levels of display. Other crime types such as Harassment or Murder reveal different patterns depending on the level of aggregation. Assault is the only type of crime that shows an increase in number of incidents over time on all levels for the entire city of Boston. Due to the representation of points it is also possible to detect smaller temporal patterns in the data. For example Larceny reveals a seasonal pattern with more crimes in summer and less crimes in winter. Furthermore, violation seems to have decreased in number between June 2015 and January 2018 but since then there might have been an increase in this type of crime again.

Now try to investigate the statistics of your preferred neighborhood(s) yourself!

Findings

To be more efficient in the fight against crime and crime prevention, a detailed analysis of the crime situation in recent years is necessary. Where do most crimes occur? What type of crimes happens where? And how has the number of crimes changed over the last years? By using our plots and maps local authorities, such as the Boston Police Department or people intending to move to a safe neighborhood with increasing security and less crimes over the past years can draw the correct conclusions to answer their questions.

By looking at specific neighborhoods instead of the big picture, it is possible to detect each neighborhood’s progress regarding goal 16 of the sustainable development goals. As for example East Boston seems to be dealing surprisingly good with a lots of crime types decreasing over time. Moreover by including demographic information in the analysis, we can better understand why some crime types happen where. Accordingly, crime prevention can be specifically adapted. While in some places larceny protection must be improved, in others investment must be made in violence prevention to reduce the crime rate further.

Regarding goal 16 of the sustainable development goals we can conclude that the city of Boston is overall on a good way in decreasing the crime rate in the city with only a few crime types showing slightly increasing numbers over some level temporal of aggregation.

Further information

Team - Contact us

Fabian Biland

Motivator

Patricia Schnidrig

Data Analyst

Fabienne Christen

Bug Fixer

Fabian Biland

heart and soul of the team (and also the quota man), keeping us motivated and on track

Patricia Schnidrig

always in for a laugh, master researcher and reason why no deadlines were missed.

Fabienne Christen

also part of the team, fixer of code and searcher of workarounds.

Sources

Data Images Literature Web Design

The design of the website was inspired by W3schools.com