Introduction

Jeremiah’s Vanishing New York is a blog run by Jeremiah Moss that reviews historically important places in New York City. I have followed his blog for about a year. He is especially skeptical and negative about replacing old/historic restaurants, cafes, and coffee shops with new, giant franchises, in regard that the city is getting lose of its identity and own culture. One of the topic he frequently mentioned is opening a new Starbucks in old, comparatively untouched neighborhood, assuming that it will result in skyrocketing rent and threatening small local business.

Starbucks is one of the largest coffeehouse chains in the world, as well as in the United States. Recently, Starbucks is considered as one that threatens local businesses. Not only local businesses, but also property values are affected as well. The property value of homes within a quarter mile of a Starbucks rose 96%, according to Zillow, between 1997 and 2013. In the same time frame, average increase for all U.S. homes is 65%. [1]

As of July 2017, there are 333 Starbucks store is in New York City only, and 2/3 of them are in Manhattan area. Starbucks used to be relatively scarce like inner Brooklyn neighborhood or Bronx, but stores are expanded to the areas that used to prevailed by local coffeehouses.

Using Carto, I visualized Starbucks stores in New York City by zip code in New York City in various ways, to see how Starbucks are distributed in New York City, which area has the highest density with the stores, and how Starbucks expands its stores geographically within the city.

Visualization Example

I referred a cartographic guide to Starbucks’ global domination from Quartz as my visualization inspiration. This article presents various maps with the worldwide Starbucks location data.

The map below shows every Starbucks in America. Each dot represents one Starbucks store. In other words, the gray areas are untouched by Starbucks yet. Each dot size is very small and has no boundary around the dot, so it looks like a heat map.

Another interesting map is “Every Starbucks on Manhattan” shown in below. This map helps to understand how the stores are distributed in Manhattan.

The author also shows the distribution in the same scale sizes. The author says the distribution in many cities shapes like each city, or at least its wealthier neighborhoods.

Data Preparation

Socrata.com provides all Starbucks location data in the world. I exported the file into Spreadsheet, and filtered by Country (US) – Country Subdivision (NY) first. There is a ‘city’ column within the data, but some of the stores are not under “New York”. For example, Jamaica is in Queens, but the stores in Jamaica is under “Jamaica”, not Queens. Therefore I needed to find the names of those neighborhoods in New York City 5 boroughs, so that any of the data is not excluded. I searched all zip codes in New York City, and used as a filter to select stores only within New York City boundary.

I also would like to find opening dates of the stores, but no organized format exists. I could obtain the opening dates for only around 100+ stores in New York City from this website.

New York City zip code boundary shape file is retrieved from NYC open data.

 

Results

After importing the csv file that contains Starbucks location data and .shp files of New York City zip code boundaries, I added  some analysis to calculate the number of Starbucks on each zip code.

On the map below, each green dot represents the location of each Starbucks. The choropleth map represents the number of Starbucks on each zip code. I added a widget on the right side of the map to show how many stores in each “area”. This data comes from a column named “city” in the csv file. As I mentioned above, the “city” does not correctly categorize 5 boroughs. In fact, “New York” indicates “Manhattan” only. There is no “Queens” in the column. Instead, Queens’ neighborhoods are separated as Jamaica, Astoria, etc. To deliver clearer message to users, the “city” column in the original data should be reorganized if further study is needed.

As users click the name, the map shows the borough with the stores.

(Click the image to the Carto map link)

 

 

This choropleth map shows the density of Starbucks per zip code. Manhattan midtown has definitely the highest store density.

(Click the image to the Carto map link)

 

I also would like to visualize how Starbucks stores geographically expanded by year, but I could find opening dates of about 100 stores only. The darker the color is, the most recent the opening date is. Since I do not have sufficient data (opening dates) of all stores, I am not able to conclude the trend how the stores have expanded geographically. In the future research, I would show dynamic changes of geographical expansion.

(Click the image to the Carto map link)

 

Article reviews 

• Starbucks Plans New Location On Border Of Bed-Stuy & Bushwick (http://gothamist.com/2017/06/28/bed_stuy_bushwick_starbucks.php)
• Tracking the Hyper-Gentrification of New York, One Lost Knish Place at a Time (https://www.nytimes.com/2017/09/27/books/review/vanishing-new-york-jeremiah-moss.html?_r=0)

Reference:

[1] http://time.com/money/4763722/property-value-starbucks-whole-foods-trader-joes/