Title
Density of Sex Offenders Surrounding Schools in
Neighborhoods of Varying Income Level in |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Author
Contact Information: fesurratt@yahoo.com |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Abstract
To show whether or not schools in lower income neighborhoods have a higher density of sex offenders nearby than schools in higher income neighborhoods. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Introduction
Initially, for my project, I wanted to do something
where I analyzed the number of sex offenders in the neighborhood surrounding
my children’s school. However, when I went to the Megan’s Law website and did
a search on their school, I found that there were very few offenders in the
area. While this was good news to me as far as the school is concerned, it
didn’t give me a very interesting project. In trying to figure out how to use
this information, I realized that I live in an upper-middle class, suburban
neighborhood and perhaps that was the reason for the low density of sex
offenders. This led me to my final project idea to analyze whether or not the
income level of the neighborhood in where a school is has any affect on the
density of surrounding sex offenders. I also chose to change my focus to |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Background
For more than 50 years, A new law that passed in 2004 provided the public with internet
access to detailed information on registered sex offenders. This expanded
access allows the public to use their personal computers to view information
on sex offenders required to register with local law enforcement under The search engine of the website allows searches by name, address, city, ZIP code, county, schools and parks. Once you get your search result window, you can click on the blue boxes (the offenders’ locations) and you get another screen that will give you a picture, physical description, address and the offense(s) for that particular offender.
The law is not intended to punish the offender and specifically prohibits using the information to harass or commit any crime against an offender. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Methods
My first decision was to determine which geographic area
to cover. As stated in my introduction, I ultimately decided on I then decided to limit my study to public, elementary schools. I chose that because that is the type of school that my children attend. Also, I feel that younger children are more susceptible to “stranger danger,” i.e.: more likely to fall for the “I lost my dog. I’ll give you candy if you help me look for him” trick than say a high schooler would. Finally, I had to decide what kind of census data to use to determine income level. I was originally going to use median household income, but realized that a single person making $50,000 annually is in a different category than a family of 5 with the same income. Therefore, I ultimately decided to go with per capita income. After I made these initial decisions, the next step was to actually find the data sets that I needed. The first step was to locate the census data. I had a little bit of trouble finding the files in the correct format but I ultimately found what I needed at http://factfinder. census.gov. Here I was able to download a table of all the census tracts with the annual per capita income. The next step was downloading a shapefile of the Now that I had set up the general geography, the next
step was to add point files of the schools and sex offenders. Finding the
data for the I was then able to geocode the schools against the
street centerline file I found for
Once I chose the schools to analyze, I needed to collect
the sex offenders’ information. Using |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Results
After all the downloading, joining and geocoding, I ended up with a dot-density map showing my chosen schools and the nearby offenders.
This map does show the number of offenders around each school. However, because of the large size of the county and the small radius around each school, it is nearly impossible to enlarge it to a size where it can be more easily interpreted. Therefore, I created a summary table to show a more accurate count of offenders around each school. Also, the chart takes into account any offenders’ addresses that were unmatched in the geocoding process while the map does not.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Analysis and Conclusion
I would say that in general, my findings are consistent with my hypothesis. They don’t show an absolute correlation, but definitely a general trend. For instance, James R. Cowan is a school in a high income level neighborhood that is relatively small and surrounded by middle and even one low income neighborhoods. In contrast, Valley Oaks is a low income school that is in a less densely populated tract. However, as you scan down the chart, you will see that the low income schools are clustered more towards the top, the middle income in the middle and the high income towards the bottom. I did as much as I could to investigate my hypothesis, however, I was limited to the resources and time available to me. As a result of these limitations, there are several factors that could potentially change (or perhaps further strengthen) my findings: 1. Sample size I only chose 28 out of over 175 schools. Would my findings be different if I had time to analyze a larger sample? 2. Outdated data ~The census data is from 2000, which makes it 8 years old. How has this information changed in that time and how could it possibly affect my results? ~All the Megan’s Law offender information comes with a disclaimer that the offender may have moved. Since an offender that moves has until 5 days after their birthdate to report their move, it’s possible that a number of offenders have moved either away from, or for that matter to, the school zones I analyzed. I also came across a handful of offenders that were tagged as being in violation of the registry. 3. Erroneous data
from Megan’s Law There were a few addresses that didn’t make sense, such as
a single address reported to be in North Highlands when all surrounding
offenders are listed as being in 4. Choice of buffer
zone If I had chosen a larger or smaller buffer zone, would that further confirm, or refute my hypothesis? 5. Some schools not
mapped on Megan’s Law A few of the schools that I had initially chosen came up with a “search not found” message at Megan’s Law. When I searched by address, a few were found, but still a couple were not so I had to swap out for another school. 6. Population
density as a factor From my dot-density map, it seems as if population density could possibly play a role in the density of sex offenders neighboring a school. It would take much more detailed analysis to determine whether income level or population density plays a greater role in my final findings.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
References and Resources
TIGER files for census tract and school district shapefiles Census demographic information Sex offender information Search engine for schools www.msa.saccounty.net/gis/default.aspx |
