Have craft breweries followed or led gentrification in Portland, Oregon? An investigation of retail and neighbourhood change

ABSTRACT

This paper quantitatively measures the relationship between retail change and residential gentrification by examining the connection between gentrification and the opening of craft breweries in Portland, Oregon. Our findings indicate that: 1) craft breweries were slightly more likely to open in gentrified/gentrifying neighbourhoods than not and, 2) with the exception of the 1990s, breweries in gentrifying neighbourhoods most often opened followed the onset of neighbourhood upgrading. During the 1990s, we find that breweries were at the leading edge of gentrification, representing the type of economic and cultural changes in the commercial landscape that make a neighbourhood attractive to middle-class gentrifiers. Breweries opening during the 1980s, 2000s, and 2010–2015 periods were more likely to follow gentrification or to open in stable or upgrading neighbourhoods. Considering these results, we caution that while craft breweries may retain local industrial jobs and revitalize commercial districts, they may also solidify on-going patterns of gentrification.

KEYWORDS:

• Craft brewing industry
• gentrification
• neighbourhood change
• Portland Oregon
• urban retail

Acknowledgement

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Disclosure statement

The authors declare that there is no conflict of interest.

Geolocation information

Portland, OR, United States of America 45.512794, -122.679565

Notes

1 The Brewers Association defines a craft brewery as one that produces less than 6 million barrels of beer annually, that has less than 25% of its ownership (or equivalent economic interest) in the hands of an alcoholic beverage industry member that is not itself a craft brewer, and that has the majority of its total beverage alcohol volume in beers whose flavour derives from traditional or innovative brewing ingredients and their fermentation (see Brewers Association 2015).

2 We do not attempt to explain the success of the craft beer industry in Portland in this paper. Common explanations in the literature include its historical strength in the brewing industry, the availability of low-cost industrial space, proximity to natural resources, historically favourable taxation laws, and a strong pub culture (Dunlop 2013; Cesafsky 2010). An anonymous review of the paper rightly questioned why Portland has seen a rise in craft breweries with gentrification, while other gentrified/gentrifying cities have not. This is an interesting question, but is outside the scope of the paper. Future comparative research could examine differences in local economies to answer this question.

3 See Spatial Structures in the Social Sciences (2014, p. 24) for details on classification of professional employees. Generally speaking, this category includes professional and management workers.

4 Other methods we considered included: 1) a spatial hedonic price model, which we deemed inappropriate because of the prohibitively high costs of home sale transaction data in our study area for a sufficient time period, because we are less interested in the amenity value of homes being located near breweries and more in the role of breweries in driving or following neighbourhood change, and because this method is less capable of capturing change over time; 2) the cross-sectional field survey method combined with census data used by Hammel and Wyly (1996), which was not selected because the large scale of fieldwork was prohibitively expensive and because we are interested in the historical relationship between gentrification and brewery presence; 3) longitudinal regression was rejected because it was unclear how this method would help answer our research question. If we used home values as the dependent variable, we would be building a hedonic price model that has the problems discussed above. If we used a logistic model to predict brewery presence or a Poisson or negative binomial model to predict the number of breweries per tract, we would run into problems due to the overall low number of breweries in the sample, especially in earlier years.

5 Note that the 2000s do not have a leading group and 2010–2015 does not have either a concurrent or leading group because we did not include census data from after 2010 in our PCA. Therefore, it is also possible that some of the non-gentrifying tracts in these two time periods may be classified as gentrifying based on future census data.

6 Additionally, the results of the seven variable PCA were similar to the four variable solution, with the significant limitation that there were more missing values in the former results due to pairwise exclusion.

7 This motivation also led us to exclude race as a variable in our PCA. While gentrification in most US cities has a race as well as class component, and this certainly holds true in Portland (Drew 2012; Gibson 2007; Goodling, Green, and McClintock 2015; Sullivan and Shaw 2011), we found that the timing of changes in the racial composition of neighbourhoods justified excluding it from our four variable PCA. A key factor here is the historical residential segregation patterns of Portland. While significant long-term residential racial segregation is present, contemporary trends in the city see overall increases in the percent non-white population. However, independent samples Mann-Whitney U tests (results available from authors upon request) show that while tracts with breweries follow the city-wide trend of increasing percentages of people of colour and are not significantly different from the rest of the city in any decade, they are increasing at a much lower rate (p < 0.01 in 2000s and 2000–2015). Indeed, the breweries opening in 2010–2015 fell in the city’s minority of tracts that are becoming whiter. Future research could build logistic regression models by decade or longitudinally using similar variables to unpack the effect of race or qualitatively explore the relationship between craft brewing and race (see also Bland 2013).

Additional information

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.