Crime and demography in Colorado’s counties
Can one draw fascinating insights about Colorado, just by solely looking at group-by aggregations of means and sums? I confidently say yes — they do say less is more. As someone coming from California, maybe my findings are already well-known and not full of splendor, but I truly enjoyed getting to learn a bit more about the state in which I go to college.
To start, what shocked me at first was that Colorado has 64 counties! To put it into perspective, California just has 58 and reigns as one of the largest states. Arizona, which is slightly bigger than Colorado, just has 15 counties. So, this was truly something to get used to in my data analysis as I’ve never heard of most of the Colorado counties I came across. Using pandas to aggregate and find the mean total population from 2008 to 2020, El Paso, Denver, Arapahoe, Jefferson, and Adams counties, in this order, have the largest populations. Those with the lowest appear to be Kiowa, Jackson, Hinsdale, Mineral, and San Juan counties in this order as well.
Furthermore, one of my favorite aspects of my project was taking a glance at the crime statistics for Colorado. What first caught my eye was when I was performing temporal data analysis, and was taking a look at the months that certain crimes occur most frequently. The crimes in the Colorado county crime dataset I used were “Crimes Against Person”, “Crimes Against Property”, “Drug Equipment Violations”, and “Drug/Narcotic Violations”. Unbeknownst to me, since 2008 to 2021, the highest average of drug violations has occurred in March (37.43) and then April (36.95) across all Colorado counties, perhaps related to springtime and spring break activities. Neatly enough, December has the lowest average of 29.42 drug violations, and also 21.40 drug equipment violations over 13 years. MI hypothesize that when people come home for the winter season and the holidays, there is less time to be out and about, and perhaps being around relatives and parents creates less of an environment that leads to drug use. June (31.23) and July (32.39) also have the lowest means right after December, and perhaps this could relate to summer break and kids going back home to their parent or family to stay.
The county with the highest average of mean monthly drug violations from 2008 to 2021 is Adams County at 277.49, surprisingly surpassing its counterparts that range in much higher populations. Denver follows shortly after but quite lower, at an average of 201.16. Jefferson and El Paso are tied around a mean of 147, and Larimer County follows at a 100.12 average. Arapahoe County, despite having the third largest population, is in 10th place with an average of 59.61 violations. For “Crimes Against Person”, the sorted averages follow a very similar order, with Adams County taking the lead again with a mean of 763.07 offenses every month from 2008–2020. Denver follows suit, then El Paso and Jefferson, and then Weld County. It is evident that Adams County, again just 5th in largest population, takes the cake with crime.
Once the pandemic infiltrated all corners of life, nearly all these averages of crime decreased. For example, drug violations went from 39.35 average amount of cases per month per county in 2019 to 29.13 in 2020. Average crimes against property dropped to 100.29 cases, originally having an average of 111.03 in 2019, having the lowest mean of any year starting from 2008. Intriguingly, “Crimes Against Person” did not drop during the pandemic. This could be a result of domestic violence increasing once everyone was subjected to their homes, as is suggested in this study by Piquero et al.
Furthermore, there are a few other interesting yet un-categorizable highlights I captured that I would like to share. First, household size has barely changed and has remained at an average of 2.3 people per household since 2010 per county in Colorado. Based on the U.S. Census of 2021, the average family/household size consisted of a larger 3.13 individuals in the U.S., so perhaps it can be inferred that Colorado families do not have children as often as other states. Additionally, when taking a look at CDC death demographics, I conducted specific analysis on Colorado deaths, diseases, and how the week of the year may play a role. I gathered the sums (not means due to NaNs) of all deaths in Colorado since 2014, and Colorado seems to be pretty in sync with “All Cause”, or the total of all U.S. deaths — the most pass during the first 10 weeks of the year, starting normally at week 1 or 2.
Unlike “Diseases of the Heart”, Colorado does not follow the typical U.S. pattern where most die in the first 10 weeks but has weeks 12, 16, and 15 around the 3rd-5th place when sorted in descending order, which accounts for March and April. This study by Samantha E. Jacobs et al discusses how springtime brings a mini-surge in the common cold and flu, which may provide reasoning for this spike. Colorado is a relatively fit and active state, leaving March and April to be the time when spring comes and it is time to venture out again. Unfortunately, the weather is still cold, and mixed with increased socializing this can easily lead to feeling under the weather. Sickness can inflame the body and put stress on the heart and lungs, and therefore speed up any of the ill effects diseases of the heart may bring, including death.
Insightful findings can be discovered even through simple group by aggregation in Pandas. Through just a couple of data sets and calculations of means and sums, there is enough information to understand Colorado counties, crime during COVID-19, Colorado deaths in comparison to the rest of the U.S., and more.
