Cafés and Gentrification

This is a study of how cafés correlate to the gentrification of neighborhoods. I plan on analyizing data in the Boise Metro Area, Des Moines Metro Area, and Richmond Metro Area. I figure that these cities are urban enough to show possible gentrification of neighborhoods, but not too large that they will be overwhelming with data. These metro areas also cover the west coast, midwest, and east coast, so the study will cover different parts of the United States.

First I downloaded a shapefile from Esri that broke the United States up into their zip codes. The zip codes will serve as the neighborhoods. Next I downloaded open street map data from Geofabrik. I downloaded shapefiles from Idaho, Iowa, and Virginia. I utilized the point of interest shapefile, and I performed a query to show only cafés. Next came the data for gentrification. There are many things one can use to measure gentrification; I decided to use housing data because it is more spatially representative. At first I tried to utilize data from zillow, but I came across a lot of problems. The zillow data was monthly as opposed to annually, and was very incomplete with how many zip codes it had data for. Even in urban areas there were gaping holes. I finally came across the Federal Housing Finance Agency (FHFA), and was able to download an excel spreadsheet for all of the zip codes and years that I needed.

First, I created a geodatabase for each metro area.To make my maps I zoomed in to the metro area on the Esri map to see which zip codes were a part of the metro area. After selecting the records for these zip codes in the attribute table, I exported the selected data to its respective geodatabase. Next, I used the geoprocessing tool to clip the café data to the metro map. I saved the new layer of clipped cafés in the geodatabase as well. I created separate excel spreadsheets for each metro area. The spreadsheets included the zipcode, annual Housing Price Index (HPI) from 2009-2019, and the percentage of change in housing price from 2009-2014 and from 2014-2019. I made a new map for each metro area, which included the metro area with zips as a base, and the cafés as a layer. I joined the metro area attribute table with the HPI spreadsheet, using the zip codes as the joining feature.

The first cafés were introduced by the Ottoman Empire. Neighborhood cafés were a place for patrons to socialize, and discuss music, art, literature, and politics. These cafés were so popular, that the government funded their own coffeehouses. The Empire established their coffeehouses in the cities they conquered (Kilkenny). These cafés were opulent places of grandeur where people could relax while drinking coffee.

Traders introduced coffee to the Europeans. The British coffeehouse "... first opened in 1652, and, by 1700, there were more than 2,000 coffee houses in London alone..." (Kilkenny). These cafés were called 'penny universities' due to the intellectual discourse that would occur for the price of a cup of coffee. Although a penny might seem small, it was actually quite expensive for the time. Only the affluent had the funds and the time to spare, to spend hours at these coffeehouses.

From Britain, coffeehouses made their way to the American colonies. After the Boston Tea Party, in which colonists acted against unfair tea taxation, "... the Continental Congress passes a resolution against tea drinking stateside" (Kilkenny). This led to a rise in the popularity of coffee in America. In the 1950's and 60's Italian immigrant neighborhoods repopularized the café. Teenagers with an excess of time and money would frequent these coffee shops "... and hang out with counterculture artists , writers, and musicians" (Kilkenny).

In the 1990's Starbucks revitalized the coffee craze. They introduced the middle class to the café lingo and gave consumers a new pastime. Although the coffee shops had been mainstreamed, the prices and the atmosphere is still reminiscent of its historical context.

Modern cafés tend to be located in affluent neighborhoods. Although there are many factors that go into the wealth and desirability of an area, the researchers Edward Glaeser, Hyunjin Kim, and Michael Luca found that the establishment of new business, such as cafés and yoga studios, tend to precede gentrification. When asking the question of whether a cafe might be the result of an increased wealth of an area rather than the cause of it, the researchers found that the "... entry of Starbucks (and cafés more generally) into a neighborhood is in fact indicative of housing price growth across the United States" (Glaeser, Kim, and Luca).

The Glaeser, Kim, and Luca state that a neighborhood's business landscape reveals its wealth, and that "Gentrifying neighborhoods might also attract more upscale establishments..." Wealthy patrons are willing to spend more locally.

This storymap will analyze the correlation between cafés and the Housing Index Prices across 3 locations: the Boise metro area, the Des Moines metro area, and the Richmond metro area.

Boise, ID is a western city fast on the rise. With an increase in local tech companies and the jobs they create, Boise has seen a revitalization of their downtown and urban neighborhoods. In order to determine whether cafés have preceded this transformation of a quiet city into middle class powerhouse, I have broken up the changes in housing prices that this city has seen over the past ten years. Housing Price Index (HPI) data was acquired from the Federal Housing Finance Agency (FHFA), and categorized by zip code. In Figure A (below) you can see that there was not much of a shift in housing prices. Boise neighborhoods saw anywhere from a -5 to a +6 percent change. To see whether cafés were associated with the changes, the cafés that were opened between 2000 and 2009 are included in Figure A. As you can see there were only 2 cafés that had opened during that time frame; both were a Starbucks. This does not take into account cafés that were already in establishment since before that time period, only cafés that began their business during that period. Out of the two cafés, only one is near any growth in HPI. Growth does not appear to be correlated to the establishment of cafés in this time period, at this location.

Moving on to the following five years, Boise saw exponential growth from 2014-2019. Neighborhoods rose between 46 and 71 percent. Again, I looked at the cafés that were opened before 2014 in order to see if they preceded a change in the market. With the opening of 6 cafés from 2009-2014, it seems as if the rise in cafés might be liked to the rise in HPI. But upon looking at the spatial location (Figure B, below) you can see that none of the cafés are located in the highest growth areas.

It appears as if Boise's rise in housing price might be attributed to some other factor. But the question of correlation still remains. If café establishment was not a preceding element, it might still be a symptom of rising HPI and thus gentrification. To look at this, Figure C (below) shows HPI percentage and cafés established between 2009 and 2019. Boise has seen an overall rise in housing price and the opening of 19 cafés over the past ten years. With the most cafés starting business in the highest risen places, and the rest having opened in the mid-to-high areas, it seems as if cafés might be concurrent to changes in the housing market in Boise.

The next area we will look at is in the midwest. Having evolved from a rural area to one of the top 10 places to raise a family, it has seen a lot of growth. Figure D (below) shows a slight rise in HPI in the northern and western areas, and a slight decline elsewhere. The 14 cafés that opened between 2000 and 2009 appear to be gathered centrally and radiate out towards the growing areas.

The map showing HPI from 2014-2019 (Figure E, below) shows a very different story. The areas central and southwest Des Moines show a more dramatic increase in housing price, while the cafés that were established from 2009-2014 are still located in the western part.

When looking at a ten year map of Des Moines (Figure F, below), it does not appear that the cafés that were opened from 2009-2019 follow the same spatial pattern as the higher rising housing price areas. The higher increase in HPI neighborhoods are more central and sort of fan out. But the cafés opened during that same time period cluster centrally and radiate towards the west. There does not seem to be any correlation in cafés opened and HPI in Des Moines.

Richmond Virginia is a smaller east coast city that is steeped with history. Looking at the map of housing price index from 2009-2014 in Figure G (below), the only areas to see a slight increase in HPI are the central and northeast areas of the city. 13 cafés were established in the nine years prior to that (2000-2009), and closely follow the same spatial pattern. While the cafés are only in the lower growth areas, the pattern is still visable.

Looking at the map of 2014-2019 (Figure H), the pattern of increase in housing prices flips. A fan around the downtown to the south and east sees a more dramatic increase in HPI. While there is still increase in the other areas, it is not so large. The cafés opened prior (2009-2014) fall into the void. The 14 cafés shown do not appear to anticipate the rise in housing prices.

Richmond's ten year HPI map (Figure I, below) shows a clear spatial pattern. The cafés established during the same time period cluster in the center and spread out towards the northeast. There is one lone café that was opened in the southeast, and the neighborhood in which it is located has appreciated more significantly than the neighborhoods around it. It appears that HPI is rising along with café openings at the same time.

Looking at the maps (Figures A-I, above) it does not appear that cafés tend to precede a rise in housing prices, but instead accompany them. Looking at a graph of how many cafés were established each year in the three metro areas (Figure J, left) we can see Boise and Richmond both had a dip in the amount of new cafés opening in 2011. When we see the graph of Housing Price Index for the same years and areas (Figure K, below left), both Boise and Richmond experience that same decline in 2011.

Another similarity can be seen in 2015. There is a sharp rise in the number of cafés in all three metro areas, and a steady incline in HPI. However, both graphs are not completely similar. In 2019, Boise has a steep decline in the number of cafés that were opened that year. But 2019 also see's Boise's HPI rise dramatically. While both graphs are not completely identical, they are similar enough to see that there is a positive concurrent correlation. It easy to see that the opening of cafés is not indicative of impending gentrification, but rather a sign of gentrification in progress.

Esri. "USA ZIP Code Areas." ArcGIS.  https://www.arcgis.com/home/item.html?id=8d2012a2016e484dafaac0451f9aea24 . Accessed 9 June 2020.

Federal Finance Housing Agency. "Housing Price Index Datasets." Federal Finance Housing Agency.  https://www.fhfa.gov/DataTools/Downloads/Pages/House-Price-Index-Datasets.aspx#qexe . Accessed 22 June 2020.

Geofabrik. "United States of America." OpenStreetMap Data.  http://download.geofabrik.de/north-america/us.html . Accessed 9 June 2020.

Glaeser, Edward, Hyunjin Kim, and Michael Luca. "Nowcasting Gentrification: Using Yelp Data to Quantify Neighborhood Change." AEA Papers and Proceedings, May 2018, 108: 77-82.  https://pubs.aeaweb.org/doi/pdfplus/10.1257/pandp.20181034 . Accessed 21 June 2020.

Kilkenny, Katie. "A Brief History of the Coffee Shop As a Symbol for Gentrification." Pacific Standard, 27 July 2017.  https://psmag.com/economics/history-of-coffee-shop-as-symbol-for-gentrification . Accessed 21 June 2020.

Rascoff, Spencer. "Confirmed: Starbucks Knows the Next Hot Neighborhood Before Everyone Else Does." Quartz, 28 January 2015.  https://qz.com/334269/what-starbucks-has-done-to-american-home-values/ . Accessed 21 June 2020.

Stock Photos. Unsplash.  https://unsplash.com/s/photos/coffee . Accessed 23 June 2020.