Defining Underserved Communities:
One half of the model is devoted to the identification of underserved communities. This was done by defining the amount of children, density, and the average income of a Census Tract Block Group. Calculations were developed using the ratios below.
Ratio of Children: number of children under age 17/ number of total residents (The target demographic was identified as children of a younger developmental stage along with those under driving age) each Census Tract Block Group was then assigned a score between (0-10). A higher percentage of children yielded a higher score.
Density: number of residents per square mile of residential land within Census Tract Block Groups. Each Census Tract Block Group was then assigned a score between (0-10). A higher density yielded a higher score.
Poverty levels were calculated by normalizing county data according to the federal standards for groups that qualified for both free and reduced meals. This was a multi-step process that included:
- Calculation of a County’s income as a percentage of the national average ($52,029) income (e.g. Baltimore City’s income ($40,087) is 77% of – or 23% lower than – the national average)
- Defining a new county poverty level based on the county’s percentage of the national average (e.g. Baltimore City, with 77% of the national average, would set the poverty line at $16,551)
- Calculation of the ratio of households at or below 185% of the county’s poverty line, based on the US Department of Agriculture’s target for families of four eligible for free and reduced meals (e.g. for Baltimore City, ($16,551)(1.85) = $31,429)
Note: This is not actually how school meal programs are calculated. They are done by federal numbers and do not vary regionally.
Each Census Tract Block Group was then assigned a score between (0-10). A higher percentage of households at or under the 185% threshold yielded a higher score.
Defining Access to Parks:
Access to park space was evaluated on the proximity to state, regional, and local parks and trailheads for public trails. This data was built from the Department of Natural Resources public lands geographic information systems (GIS) layer, assembled data from ADC maps, and local data where available.
Proximity to park space included those parklands outside of Census Tract Block Group and county boundaries. It was calculated as a ratio of mean distance to park/ size of Census Tract Block Group area to account for the large size of more rural places.
Each Census Tract Block Group was then assigned a score between (0-10). A further distance from the park yielded a higher score.
Combined Score:
The above factors were scored and then added together to produce a combined park equity score. Every Census Tract Block Group in the state was assigned a score of (1-10) for each of the four factors. These factors were then weighted according to the model and added together for a resulting score of (0-100).
The model formula is as follows:
Combined score = Concentration of Children (0-10) + Concentration of Low Income (0-10) + Density (2 x (0-10)) + Proximity to Park Space (4 x (0-10)) x 1.25
The resulting score is a 0-100, with the higher score indicating a higher need for park space.
Final scores are displayed in a quantile manner, categorizing block groups into five categories- measured against each other.
