Introduction
Discussions of gentrification and the housing crisis have dominated urban policy discussions across the United States in 2018. Many activists argue that the profit motive is to blame, and that housing should be a human right, while others contend that developers would build cheaper housing if regulations enabled cheaper development. This project seeks to explore issues around housing, economic status, and race in Brooklyn, and whether there are any significant correlations to be found in those datasets. We'll take DOB data on Illegal Conversions and Legal Permit Applications for New Residential Buildings, and try to spot any geographic or demographic correlations, hopefully inspiring further study.
Data
Our data is sourced from a combination of municipal and federal level datasets. In order to get information about successful permit applications for new residential construction and complaints about illegal conversions of units to residential use, we turn to NYC's Open Data portal, with the former coming from the NYC Department of Buildings (DOB) Complaints Received
dataset and the latter from their Permit Issuance
dataset. That data is paired with American Community Survey (ACS) 5 Year Estimates for demographic and economic data. We use the years 2012 and 2016 for each of the datasets, in order to see what changed in the intervening time period. The analysis looks at the shifts in population percentage for White, Black, Hispanic/Latinx, and Asian New Yorkers. The Census data is clean, thanks to the diligent work of the people at the Census Bureau in packaging everything. Some of the DOB data is less so, because it depends in part on user input and forms filled out by humans-- the absence of data for some fields is common, but thankfully it shouldn't impact our analysis too much.
The Permit Issuance dataset comes with latitude and longitude already calculated, which is convenient because we'd like to represent that data as points in order to spatial join it to our census tracts. The 311 data does not come with that information prepackaged, so we are forced to get creative. We have the address data over several features, so we can concatenate that and then use the Geocoder
library for Python and send our addresses off to the ArcGIS API to be processed. From there we spatial join our points to our shape-file (census tracts) to aggregate, and then move on to analysis and visualization.
Methodology
Joining the DOB datasets to the Census data allows us to better understand how racial and economic factors across Brooklyn are related to housing provision (whether legal or illegal). First, we can look at how neighborhoods have been changing over time. The following graph shows the change in ethnic makeup of a census tract from 2012 to 2016, for White, Black, Latinx, and Asian New Yorkers.