#Introduction

Article: https://projects.fivethirtyeight.com/redlining/

The data presented derives from a study conducted on the continuation of redlining where the historical use of the HOLC grading system based on race population that determined what areas were best (graded A) or desirable (graded B) to invest in (mostly white areas), and those that were declining (graded C) or hazardous (graded D) (mostly black areas) still are prevalent in the segregation of many cities.

This data analysis looks into the segregation of the cities of upstate New York relative to their HOLC graded living areas

#Loading Packages & Reading data

library(tidyverse)

## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.3.6     ✔ purrr   0.3.4
## ✔ tibble  3.1.7     ✔ dplyr   1.0.9
## ✔ tidyr   1.2.0     ✔ stringr 1.4.0
## ✔ readr   2.1.2     ✔ forcats 0.5.1
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

library(RCurl)

## 
## Attaching package: 'RCurl'
## 
## The following object is masked from 'package:tidyr':
## 
##     complete

 x <- getURL("https://raw.githubusercontent.com/fivethirtyeight/data/master/redlining/metro-grades.csv")
 redlining_data <- read.csv(text = x)

#Creating NY Subset

ny_redlining <- redlining_data %>%
  select("metro_area", "holc_grade", "pct_white", "pct_black", "pct_hisp", 
         "pct_asian", "pct_other") %>%
  filter(grepl("NY", metro_area)) %>%
  mutate(pct_nonwhite_sum = rowSums(across(c(pct_black, pct_hisp, pct_asian, pct_other))))

#Checking for Missing Variables

sum(is.na(redlining_data))

## [1] 0

#Comparing HOLC Grade between White and Non White population percentages

df <- ny_redlining %>%
  select("metro_area", "holc_grade", "pct_white", "pct_nonwhite_sum")

unique(df$metro_area)

## [1] "Albany-Schenectady-Troy, NY"          
## [2] "Binghamton, NY"                       
## [3] "Buffalo-Cheektowaga, NY"              
## [4] "Elmira, NY"                           
## [5] "New York-Newark-Jersey City, NY-NJ-PA"
## [6] "Poughkeepsie-Newburgh-Middletown, NY" 
## [7] "Rochester, NY"                        
## [8] "Syracuse, NY"                         
## [9] "Utica-Rome, NY"

#Regions designated as grade A by HOLC

a_b <- df %>%
  filter(grepl("A|B", holc_grade)) 


print("Average Percentage of White People living in greenlined areas (A or B HOLC graded areas)")

## [1] "Average Percentage of White People living in greenlined areas (A or B HOLC graded areas)"

mean(a_b$pct_white)

## [1] 66.67889

print("average Percentage of Non-White People living in greenlined areas (A or B HOLC graded areas)")

## [1] "average Percentage of Non-White People living in greenlined areas (A or B HOLC graded areas)"

mean(a_b$pct_nonwhite_sum)

## [1] 33.31833

#Regions designated as grade C and D by HOLC

c_d <- df %>%
  filter(grepl("C|D", holc_grade)) 

print("Average Percentage of White People living in redlined areas (C or D HOLC graded areas)")

## [1] "Average Percentage of White People living in redlined areas (C or D HOLC graded areas)"

mean(c_d$pct_white)

## [1] 44.98667

print("Average of Percentage of Non-White People living in redlined (C or D HOLC graded areas)")

## [1] "Average of Percentage of Non-White People living in redlined (C or D HOLC graded areas)"

mean(c_d$pct_nonwhite_sum)

## [1] 55.01333

#Conclusion

In the 6 upstate NY cities used to investigate the continue segregation between white and non-white people into greenlined and redlined areas, it was found that the greenlined areas contained 66.68% of white people vs the 33.32% of non-white people. For the redlined areas in the investigated cities, it ws found that 44.99% of white people lived in these areas white 55.01% lived in the areas. White people are more likely to reside in areas deemed as investable while racial minorities are more likely to reside in neglected areas. To further this study, I recommend gathering average demographic information such as average income. By doing so, the segregated areas can be compared to elucidate if income differences is a clear distinction between populations.