Environmental Factors that Affect Burglary in Waterloo Region
July 2004

Abstract

There are a number of factors that affect the likelihood of a certain area to have a high occurrence of burglary. The FBI has extensively collected data and analyzed it to discover patterns. Some of the social factors that affect burglary are a neighborhood's attitude toward police and the family situations of the people living in the neighborhood. These factors can not easily be mapped or quantified, but there are geographic factors that affect burglary rates as well. My study focuses of the geographic factors that affect burglary and where these factors are most prevalent in Waterloo.

Introduction

The purpose of this project is to identify the areas in Waterloo most susceptible to burglary. My study looked at the geographic factors that affect burglary and based upon them, the areas in Waterloo most likely to experience burglary. The factors I looked at were proximity to specific types of land use (open area, commercial and industrial), average income for the neighborhood tract, population density and the percentage of young population (aged 10-24). Based on a weighted average formula, these factors can be combined to form a map of Waterloo showing various degrees of likelihood of burglary.

Background Research

There were two main places that I derived my research from. One, a study by Xiaowen Yang and Richard Schneider entitled "Identifying the effects of physical Identifying the effects of physical environment features on burglary and environment features on burglary and controlling demographic and social controlling demographic and socio-economic variables". This study focused on the burglary statistics form Gainsville, FL and based on their hypothesis of social and environmental factors determined the percentage of burglaries, which followed what they considered to be significant factors affecting burglary. They looked at land use, demographic information, economic conditions and social factors.

My second resource, was the FBI's Crime Factors website. This website lists the factors that the FBI believes affects burglary, as well as factors related to other crimes. Using these two sources are looking at the data that was available, I mapped as many of the FBI's criteria as possible to get an accurate picture of the likelihood of burglary in different parts of Waterloo.

Objectives

To map these geographic conditions of Waterloo:
- Income Data
- Buffered Land Use Information
- Percentage of Young Population
- Population Density
- Likelihood of Burglary in Various Parts of Waterloo

Methodology

Click for Full-Size Image

I obtained the following data sets:
- Ea_UTM17.shp
- Income data for Kitchener (91 areas)
- 2001 Census Data Kitchener (91 areas)
- Land_uses.shp (Waterloo)
- Population Age Data Kitchener (91 areas)

I joined the downloaded tables to the Ea_UTM17 shape file. I converted each new theme to a grid. I then buffered the industrial, commercial and open area land uses to 300m using concentric rings. The buffered land use theme was reclassified so that the values weren't from -300-300, but rather 1-6, with -300 and +300 both being the equivalent of 6.

Calculations were done to determine the people / area which was people / acre. This was then converted to people / hectare and reclassified to "degree of population density". The area was found using the area calculation tool, on the Ea_utm17.shp theme.

The income data was the easiest to map because each income was already tied into a land area in the table. I obtained the population numbers for people aged 10-24. This is a wide range, but that is all that was available from the Statistics Canada website. These numbers were divided by total populations to determine the percentage of young populations for each area in waterloo. This information was mapped and then reclassified.

After four final maps were created, the map calculator was used using a weighted average formula to determine the areas in Waterloo most susceptible to burglary.

Each them was classified to roughly the values below:
Income Data (42000 - 122000)
Land Use (2-5)
Percentage Young Pop (13-43)
Pop Density (1-9)

Weighted average formula used:
(1/15000)Income Data + Land Use + 0.1(Pct Young Pop) + 0.5(Pop Desnity)

Using this formula to add the themes together, more emphasis was put on income data and land use, while less was put on percentage of young population and population density.

Results:

Income Data

Click for Full-Size Image

This map shows the average yearly income of families in Waterloo region. This data was mapped to a cell size of 20m, as were all of the maps. Towards the outskirts of Waterloo, there are fairly constant income ranges, while in the middle there is a very diverse range.

Degree of Population Density

Click for Full-Size Image

This map was reclassified to a non-uniform scale of population density. Near the middle of Waterloo were the sample land sizes are very small, when people is divided by land, a very large number like 4000 people / hectare is calculated. I'm not sure if this is faulty data in either population or area, but something isn't right. For the most part, the population in waterloo is between 10 people / hectare around the outskirts and around 150 people / hectare near the core. Since a linear scale would have a high range of 4000 people / hectare and a low range of 100 people / hectare, this is not an accurate reflection of the real population density degree, so it was reclassified using a non-linear scale.

Land Use (Buffered)

Click for Full-Size Image

Using the land_uses.shp file in the Waterloo data set, I buffered 3 specific types of land uses. Commercial, Open Areas and Industrial. These specific land uses were chosen based on the research done. 3 concentric ring buffers, 100m apart were used with buffer overlap. Since both sides of features were buffered, this created a scale of -300-300. This theme was reclassified to a scale of 1-6, which 6 meaning that an area was within 100m of all 3 types of land use. In reality the highest value that was mapped was 5.

Percentage of Young Population

Click for Full-Size Image

This map is fairly self-explanatory. It shows the percentage of the population aged 10-24, for each area of Waterloo. This data was mapped based on the FBI Crime factors discussed earlier.

Likelihood of Burglary

Click for Full-Size Image

This is the final resultant map. It has all 4 maps combined using a weighted average calculation (shown and discussed above). The areas most susceptible to burglary are darker, while the areas less susceptible to burglary are lighter.

Discussion

Based on the geographical information alone, the dark areas shown in my final map (Likelihood of Burglary (Waterloo)), are the most likely to be burglarized in Waterloo. The darkest area is also one of the richest in Waterloo. This means that the residents might employ additional policemen to patrol the area, and I'm sure that many have high-tech alarm systems. These are factors that were not considered in the study. Also numerous social factors were not considered, such as the ethnical composition of the neighborhood. So, this is not a complete study where all factors were taken into consideration.

The final result is certainly valid to some degree, but unfortunately the burglary occurrence data is not available to validate the claim. The weighted formula I used was not based on scientific calculation, but rather my own approximations. This could also call into question the validity of the result. There is not much scientific information available about the environmental factors affecting burglary, so I had to work with what I could find.

I am not very confident in saying that more burglaries occur in the dark areas on the final map. Although I am confident in saying that based on geographic information alone, statistically, more burglaries should occur in those areas. A salesman could use that information to strengthen his sales pitch, but unfortunately not enough information was available to scientifically analyze Waterloo in terms of burglary occurrence.

The darkest area on the final map does not rank highly on every preliminary map. It was an area of high income, low population density, average rates of proximity to the buffered land uses and had a high percentage of young population. These factors all combined to from a geographic standpoint, are the most likely place in Waterloo for burglaries to occur.

Conclusion

Based on environmental factors only, the most likely area in Waterloo to be burglarized is a neighborhood in West Waterloo as shown n the Likelihood of Burglary map. This study does not take into account all of the factors that contribute to burglary rates. The objectives of this study have been met. The geographic conditions stated earlier have been mapped and a conclusion has been drawn.

Works Cited

Surprising little information is available of the environmental factors that affect burglary rates, but this is what I did find.

FBI Crime Factors (2002)
http://www.fbi.gov/ucr/cius_02/html/web/crimefactors.html

Florida Burglary Occurrence Case Study
http://www.ojp.usdoj.gov/nij/maps/boston2004/papers/Yang.pdf