Jasmine Pham
Throughout this report:
JP-Civ: Justice of the Peace Court of Civil Jurisdiction
COVID-19 Pandemic Period: The period from March 2020 to present (April 2021).
EFs: An abbreviation of Eviction Filings.
COVID-19 EFs: The EFs dated to the COVID-19 Pandemic Period.
Historical EFs represent census tract specific weekly averages of EFs between 2016 and 2019 1.
Race and Ethnicity are treated as a single group for analysis.
We use Latine as a gender-neutral term that represents individuals who identify as people of Latin American descents.
Whenever a Race/Ethnicity category is calculated from ACS 2 data, White American population estimates exclude Hispanic/Latine individuals.
All visuals here are interactive, meaning that when users hover over the data, unique identifiers for each data point will display. The legends are also “operational” for they can be utilized to include/exclude any subset of the data.
In the state of Delaware, courts oversee and operate different types of judicial processes including housing related matters. Because of this, the distance traveled by a defendant to a near (or nearest in most cases) court can serve as either a risk or protective factor depending on the circumstance at hand.
In the case of evictions in Delaware, typically landlords must visit the nearest Justice of the Peace court of Civil Jurisdiction (JP-Civ) to the unit in question in order to submit an eviction filing. Even though electronic filing (e-filing) has been made available throughout the COVID-19 pandemic, there are a number of restrictions and requirements that further complicate the usability of the e-filing process. For example, the additional technology fees associated with every document submitted can hold landlords and/or tenants back from adapting the e-filing system. With the assumption that most eviction cases are submitted in person, especially during a pandemic period where commuting might have not been as advisable, we aimed to evaluate whether the magnitude of EFs was a function of the median distance between each postal address and its nearest court location within each census tract.
To examine this hypothesis, we first obtained physical addresses of all JP-Civ courts in Delaware from the Delaware Courts and all postal addresses in Delaware from Open Addresses and Open Data ArcGIS. Then, we geocoded these addresses and utilized the R library geodist to compute the geodesic distance, or, in a colloquial term, “as the crow flies” distance, from each postal address to every Justice of the Peace court in Delaware. From here, we were able to identify the nearest courts, by linear distance, to each individual household.
It is worth mentioning that we settled with the geodesic distance because given our large sample size of addresses, (1) the API services employed to calculate the actual travel distance and time are unaffordable, and (2) at present we do not host the computing power to query sufficient API requests in a reasonable amount of time.
From these distances, we were able to identify the number of households each court serves to be as follows:
In Delaware, there are only 5 JP-Civ of civil jurisdiction, among which 3 are in New Castle County and only 1 each in Kent and Sussex. Therefore, while Sussex is not the most populous county in Delaware, the sole JP 17 here technically has to serve the highest number of households in compared to any other courts.
The distribution of distances to the nearest JP-Civ courts varies greatly by county in Delaware. In New Castle County, more than 97% of the households live within 10 miles from their nearest courts. Meanwhile, for there is only one JP-Civ in Kent and Sussex, their ranges of distances are much wider, with the longest distance reaching up to 25 and 23 miles, respectively. Unlike the right skewed distributions in New Castle and Kent, the distribution of distances to court in Sussex slightly skewed to the left. Here, more than 80% of the households must travel over 10 miles to get to the court. In Kent, there seem to be 3 underlying subgroups represented in the distribution as suggested by the noticeable peaks around 1.5, 7.5, and 15 miles of distance. This could possibly be attributed to by some unique geographical boundaries.
Overall, the distribution of distances to the nearest courts from every postal address in Delaware is highly skewed to the right. For this reason, when conducting any analysis on a census tract level analysis, we decided to move forward with evaluating the median instead of the mean distance as it captures the central tendency of the whole area better. Either way, there is a very strong correlation between median and mean distance with an R2 of 99.8%.
Given that both distribution of median distances to the nearest courts is right-skewed and that their ranges vary among counties, we failed to observe clearly the nuances in New Castle and Kent County distributions respective to the state of Delaware as a whole. As such, in order to fully capture their relative standings, we mapped the median distance in quartiles normalized to each county. Each quartile represents 25% of the distribution of median distance to the nearest courts in a county. Specifically, the lowest quartile represents the group with the shortest median distance traveled to courts, and the highest quartile the longest median distance traveled to courts. For simplicity, throughout this report, the quartiles of median distance to the nearest courts by census tract will be referred to as: Quartile 1, Quartile 2, Quartile 3, and Quartile 4.