Performed data cleaning, preparation and analysis of West Midlands district and lower layer super output area level crime and deprivation data. Tech stack utilised - Python (pandas, Matplotlib and Seaborn), APIs.
Project completed by Shannon Watts and Gary Whitney.
'Crimes of despair' (The Independent, 2020).
Are social disadvantage, poverty and crime linked?
What is the link between crime and deprivation? Are you more likely to see higher levels of crime in areas with high deprivation levels? Does the type of crime differ depending on the level of deprivation in an area?
The IMD score is created with 7 domains of deprivation: Income (22.5%), Employment (22.5%), Education (13.5%), Health (13.5%), Crime (9.3%), Barriers to Housing & Services (9.3%), and Living Environment (9.3%). We expect that because crime is one domain used to create the IMD score that the crime rates in the LLSOAs should correlate with the IMD score.
The higher the IMD score the higher the rate of crime in a LSOA within a district.
- Does the crime data sourced from the West Midlands Police database correlate with the Index of Multiple Deprivation (IMD) score?
- Which district and Lower Layer Super Output Areas (LLSOAs) have higher rates of crime?
- What type of crime might you be exposed to in certain LLSOAs?
- Does the data and analysis support the generalisation that the higher the level of deprivation the higher the crime levels in an area?
There is a weak positive correlation between LSOA Index of Multiple Deprivation score and crime count. Walsall has the strongest positive correlation and this supports our hypothesis. The other districts do not tend to have a significant majority of LSOAs that have crime counts higher than 300. Sandwell had the weakest linear correlation with higher counts of crime in all LSOAs. This highlights that not all areas with a high deprivation score have high crime counts, and vice versa, the areas with lower deprivation levels still see high levels of crime.
A lower IMD score does not mean that there will be less crime and a higher IMD score does not mean that there will be a higher crime rate.
Other interesting findings:
There are higher bike theft rates nearer the city centres, by universities, and near hospitals.
Each district in the West Midlands has a similar crime trend. Violent and sexual offences make up the largest proportion of crime for all districts. And the other types of crime also follow similar patterns.
Depending on which district and LSOA you are in you will witness different crimes, e.g. In Solihull you are more likely to witness vehicle crime or theft, whereas in Wolverhampton there are more violent offences.
Wolverhampton has the highest rate of violent and sexual offences in the West Midlands.
Using the models of regression we could potentially predict the crime levels for 2023 if we aggregated the data for 2020 and 2021. We may also find patterns.
It would be interesting to see how the crime data in the west Midlands compared to other regions.
This data might disprove certain stereotypes / discrimination for areas with higher deprivation levels. This feeds into the literature examined in the research purpose.
The Independent (2020) https://www.independent.co.uk/news/uk/crime/crime-young-people-poverty-uk-trauma-discrimination-racism-a9682881.html (accessed 22.04.22).