Investigating the relationship between affluent neighbourhoods and access to affordable and highly rated food establishments in New York City.
<Ng Yim Chew Klo’e, Klo-e, kyn227>
1. Introduction
Neighbourhoods in New York City play an important role of place-making, where various neighbourhoods have a different reputation (for example the Financial district is known for great museums and historical sites and Times Square having a reputation of congregating of theatre and entertainment of all sorts) that collectively make New York City an attractive place to be in. More importantly, neighbourhoods give residents a sense of home at a more granular scale, and are often a determinant for potential house owners when sourcing for a place to stay in. There are various ingredients that contribute to good neighborhoods, such as accessibility to public facilities such as parks and schools, low crime rate and range of retail stores provided in the vicinity. Food and Beverages (F&B) establishments are also an indicator, where popular establishments mean more footfall in the neighbourhood and more “eyes on the street”, leading to public safety. Conversely , poor urban neighborhoods have been labeled “food deserts” with few grocery stores and mainly fast food restaurants ( Schuetz et al , 2012).
In a Seminal working paper published by Glaeser, Luca and Kim (2018), yelp data was utilised to quantify neighbourhood change in cites, understand and predict gentrification.
This paper thus seeks to utilise yelp data to investigate if whether there is a relationship between median rent of households and ‘cheap and good’ F&B establishments.
2. Data
There were a couple of key sources of data used to piece the dataset together for investigation:
(1) Information about businesses – ratings (on a scale of 5), price (on a scale of 4), business location (longitude and latitude) from YELP
(2) Neighbourhood boundaries of New York City from NYC Open Data
(3) Demographic information on each Neighbourhood from NYC Planning department
(4) Neighbourhood locations
2.1. Limitations and Issues with Data:
YELP data was limited in a way that the query results will only return the first 1000 results (of food establishments in New York City) which was too little for my analysis. I went around this problem by locating the centroids of neighbourhoods and ran the code such that i will get the YELP of each individual neighbourhood instead of the whole New York City. Additionally, neighbourhoods are defined rather differently across entities (such as YELP, AirBnB as well as the New York Planning Department). While i was able to eventually match the neighbourhoods between NYC planning and YELP, there was inevitably an eventual loss in data set (lesser neighbourhoods).