Article

Urban Studies
1–19
Ó Urban Studies Journal Limited 2016
Gentrification, displacement and the Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
arts: Untangling the relationship DOI: 10.1177/0042098016680169
usj.sagepub.com
between arts industries and place
change

Carl Grodach
Queensland University of Technology, Australia

Nicole Foster
University of Texas at Arlington, USA

James Murdoch
2M Research Services, LLC, USA

Abstract
The arts have long played a role in debates around gentrification and displacement, yet their roles
and impacts as change agents are not clear-cut. According to the standard account, artists facili-
tate gentrification and ultimately engender the displacement of lower income households, but
more recent research complicates the accepted narrative. This article seeks to untangle the rela-
tionship between the arts, gentrification and displacement through a statistical study of
neighbourhood-level arts industry activity within large US regions. The findings indicate that the
standard arts-led gentrification narrative is too generalised or simply no longer applicable to con-
temporary arts-gentrification processes. Rather, the arts have multiple, even conflicting relation-
ships with gentrification and displacement that depend on context and type of art. These results
have important implications for how we study the role of the arts in neighbourhood change and
for how governments approach the arts and creative industries in urban policy.

Keywords
arts, creative economy, cultural industries, displacement, gentrification

Received March 2016; accepted October 2016

This article seeks to untangle the relationship

between the arts, gentrification and displace- Corresponding author:
Carl Grodach, Queensland University of Technology,
ment through a study of arts industries in
Property and Planning Discipline, GPO Box 2434, Brisbane,
large US regions. Most literature concentrates Queensland 4001, Australia.
on gentrification in relation to individual Email: [email protected]

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

2 Urban Studies

artists and their residential choices. Much less 2011). Yet, many argue that while gentrifica-
is known about how artistic businesses and tion deconcentrates poverty in specific
establishments relate to gentrification and dis- places, it does not occur without significant
placement. It has long been understood that social tensions and the displacement of
an arts presence helps to facilitate gentrifica- lower income households and businesses
tion – the process by which neighbourhoods (Bridge et al., 2011; Chaskin and Joseph,
suffering from disinvestment experience an 2013). Gentrification may exacerbate eco-
influx of capital and of middle and upper class nomic inequality and destroy community
residents (Ley, 1996; Smith, 1996; Zukin, ties while reducing the stock of affordable
1982). The common narrative is that artists housing at large (Lees et al., 2008; Newman
move to a neighbourhood perceived as and Wyly, 2006; Slater, 2006).
blighted and set the stage for gentrification by For these reasons, gentrification has
renovating and upgrading aging industrial, remained a central debate in urban studies.
residential and commercial buildings. Their The arts have played a key part in that
efforts serve to change the look and feel of debate yet their roles and impacts as change
urban neighbourhoods, which attract higher agents are not clear-cut. The standard arts-
income groups that often displace long-time led gentrification narrative is too generalised
residents and businesses as well as the artists or simply no longer applicable to contempo-
themselves (Curran, 2010; Deutsche and Ryan, rary arts-gentrification processes. As recent
1984; Ley, 2003; Lloyd, 2010; Mathews, 2010; research shows, in certain contexts, artistic
Pratt, 2009; Zukin, 1982, 2010). The arts may activities may generate neighbourhood revi-
catalyse gentrification through public policy as talisation without displacement and to the
well. Increasingly, cities turn towards invest- benefit of existing residents (Foster et al.,
ment in arts institutions and cultural districts, 2016; Grodach, 2011; Markusen and
and the attraction of creative industries, Gadwa, 2010; Stern and Seifert, 2010). Other
exploiting their cultural capital to generate work demonstrates that the standard rela-
development interest and consumer spending tionship may be reversed – artistic businesses
(Cameron and Coaffee, 2005; Catungal et al., and organisations may gravitate to areas
2009; Grodach, 2010, 2012; Hutton, 2009; Ley, with pre-existing creative industry clusters
2003; Mathews, 2014; Ponzini and Rossi, 2010; and affluent, professional populations
Sasajima, 2013; Zukin and Braslow, 2011). (Murdoch et al., 2016; Schuetz, 2014;
Each of these familiar types of arts-led Schuetz and Green, 2014). In other words,
gentrification remains embroiled in the artists and arts-related businesses may actu-
wider debate surrounding gentrification and ally seek out gentrified areas where their
place change. Gentrification has the poten- patrons work and live. Moreover, different
tial to bring benefits to residents in disad- forms of artistic activity (i.e. artists, arts
vantaged urban neighbourhoods if it occurs organisations, artistic businesses) may have
without widespread displacement. As gentri- different relationships to neighbourhood
fication deconcentrates poverty, it may change so that while some forms of art may
enhance access to quality housing, education be associated with gentrification (and poten-
and employment (Freeman, 2005; Freeman tially displacement), others are not (Grodach
and Braconi, 2004). In conjunction, the et al., 2014). Indeed, most work has focused
influx of higher income households is associ- on how individual artists influence gentrifi-
ated with new investments and amenities, cation rather than the effect of artistic busi-
higher real estate values and lower crime nesses and establishments, which are of
(Brown-Saracino, 2010; Papachristos et al., increasing economic importance to many

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 3

regions and a target of urban policy. we study the role of the arts in neighbour-
Further, these relationships may vary by hood change and for how governments
urban and regional context. Gentrification is approach the arts and creative industries in
not only more typical in some places than urban policy.
others, but has occurred at different periods
of time in different cities. New York has
experienced earlier, broader waves of gentri- Unpacking the arts-gentrification
fication than Dallas and, to a lesser extent, relationship
Chicago for example. Scholars recognise that gentrification has
Given the potential for different and con- taken on different forms and proceeds at a
flicting outcomes, and the dearth of research differing pace and scale across time and
on art industries compared to that on artist place, often with successively higher levels of
residence, this article examines the relation- capital and corporate and state investment
ship between an arts industry presence, gen- (Hackworth and Smith, 2001; Smith, 1996).
trification and displacement. We ask the At each stage in the ‘life cycle’ of gentrifica-
following questions: How is the presence of tion, artists, arts organisations and artistic
arts establishments associated with gentrifi- businesses are framed as playing a central
cation and displacement? Alternatively, do role as agents and supporting actors in
gentrifying or gentrified areas attract an arts neighbourhood change (Cameron and
industry? Do the arts avoid gentrified places Coafee, 2005; Deutsche and Ryan, 1984;
altogether? To address these questions, the Ley, 2003; Lloyd, 2010; Mathews, 2010;
study statistically examines the relationship Yoon and Currid-Halkett, 2015; Zukin,
of arts industries to gentrification and dis- 1982; Zukin and Braslow, 2011).
placement in different urban and regional In the most common account, the gentri-
contexts. Using data on arts establishments, fication process is initiated as educated indi-
we examine two types of art activity, the fine viduals with modest incomes are attracted to
arts and commercial arts, in US metros with the ‘gritty authenticity’ and low-cost func-
a population over 2,000,000 between 2000 tionality of aging industrial buildings and
and 2013. We also hone in on the patterns in dilapidated central city neighbourhoods
four large regions with different arts indus- (Glass, 1964; Ley, 1996, 2003; Lloyd, 2010;
try characteristics: Chicago, Dallas, Los Silver and Miller, 2013; Zukin 1982, 2010).
Angeles and New York. The findings indi- In this context, artistic activity is considered
cate that the arts-gentrification narrative to spark the first stage of gentrification by
does not hold up as a generalised pattern engendering the early transformation of a
across different types of arts activity. We disinvested urban area. Since Zukin’s (1982)
find that arts industries generally do not study of gentrification in SoHo, many have
play a significant role in gentrification and observed that artists are attracted to such
displacement. In fact, arts industry growth is places based on their affordability and
weakest in gentrifying areas and it is gentrifi- adaptability and the aesthetics of declining
cation that predicts arts growth in most con- historic and industrial buildings and neigh-
texts. The exception to this occurs only in bourhoods (Chang, 2016; Deutsche and
those neighbourhoods with a concentration Ryan, 1984; Ley, 1996, 2003; Mathews,
of arts activity that are located in regions 2010; Shaw, 2013). As artists and arts orga-
where the arts are not widely concentrated nisations make these neighbourhoods home,
and gentrification is not widespread. These they establish a ‘bohemian’ atmosphere and
results have important implications for how create new economic value through small-

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

4 Urban Studies

scale artistic businesses and neighbourhood In each of these scenarios, the arts –
amenities including bars and cafes (Hutton, whether clusters of artists, artistic busi-
2009; Lloyd, 2010; Zukin and Braslow, 2011). nesses, arts facilities or arts districts – are
The physical and symbolic appropriation framed as changing the character of an area
of a disinvested neighbourhood by these and, in this way, are assumed to play a cau-
‘pioneer’ gentrifiers changes area character sal role in the gentrification process. While
and sets the stage for future rounds of gen- this research provides important insights
trification (Brown-Saracino, 2010; Cameron into how the arts can facilitate place change,
and Coafee, 2005; Ley, 1996, 2003; Smith, it does not tell us if gentrification and displa-
1996; Zukin, 1982). Here, a new wave of cement would have occurred without an arts
migration occurs. Spurning the homogeneity presence or the extent to which the arts gen-
of suburban living, professionals with higher erated change in relation to other factors.
incomes settle in the neighbourhood in While case studies document specific scenar-
search of ‘authentic’ urban places (Ley, ios in particular places, they are not able to
1996; Zukin 1982, 2010). Newcomers are establish if artistic activity in fact causes gen-
more likely to purchase and renovate prop- trification and displacement or if the arts
erties, which begins to price out former resi- happen to locate in neighbourhoods that are
dents and first-wave gentrifiers, including already attractive to higher income groups
artists and small arts businesses. These and investors for other reasons. Without
groups in turn are forced to seek out accom- studies of neighbourhoods with and without
modations elsewhere in the city, often in an arts presence or those that control for
adjacent neighbourhoods, where the gentrifi- other influential factors, we do not know if
cation process begins anew (Shaw, 2008). the arts actually cause gentrification and dis-
Finally, this lays the foundation for placement or if they play other roles in
broader and deeper levels of capital accumu- neighbourhood change.
lation (Hackworth and Smith, 2001; Lees, Another gap in the literature is that most
2003; Wyly and Hammel, 1999). The initial case studies focus on gentrification in rela-
stages of gentrification pave the way to fur- tion to artist residence rather commercial
ther upscale urban neighbourhoods and are arts activity and arts facilities. Given the
accompanied by additional displacement. varying types of gentrification and arts activ-
Developers, property investors and govern- ity (and the increase of arts-based economic
ments spearhead new redevelopment proj- development strategies), it is likely that dif-
ects marketed to affluent buyers that ferent forms of arts activity play multiple,
incorporate luxury residential high-rises with even conflicting roles in gentrification and
corporate and boutique retail. These projects displacement within and between places. As
often include flagship museums, performing recent research highlights, different arts
arts venues or smaller-scale arts activities to activities exhibit different relationships to
catalyse development and attract the interest neighbourhood change and some arts activ-
of creative class consumers. Similarly, they ity may be attracted to gentrifying or already
may brand areas as arts districts and partner affluent areas. As Grodach et al. (2014)
with arts organisations to promote develop- show, while commercial arts industries are
ment (Ashley, 2014; Catungal et al., 2009; associated with rapid gentrification, the fine
Cameron and Coafee, 2005; Grodach, 2010, arts avoid such places and seek out slower
2012, 2013; Mathews, 2014; Ponzini and growth neighbourhoods with few signs of
Rossi, 2010; Sasajima, 2013; Zukin and gentrification. In New York City, Schuetz
Braslow, 2011). (2014) and Schuetz and Green (2014) find

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 5

that galleries tend to locate in affluent areas zip code tabulation areas (ZCTAs) to
with expensive housing and high levels of approximate the geographical areas each zip
amenities while Murdoch et al. (2016) code route corresponds to. Thus, in addition
demonstrate similar results for nonprofit to the arts data discussed below, we can eas-
arts organisations. The evidence is also not ily link census data to our sample of zip
clear as to if and how the arts are actually codes. We use the 2000–2013 study period
connected to displacement. Research by because it allows us to capture recent signifi-
Stern and Seifert (2010) shows that arts clus- cant rounds of property development before
ters in neighbourhoods with high levels of and after the Great Recession. The limita-
both poverty and professionals exhibit lower tion of this time period is that it does not
poverty rates and increased housing values allow us to account for earlier waves of gen-
over time without displacement. Foster et al. trification or gentrification that occurs over
(2016) similarly find that new arts organisa- a longer time period.
tions temper neighbourhood disadvantage.
Using industry data on arts establishments, Variable creation
we test these possible arts-gentrification- To measure gentrification and displacement,
displacement relationships in a large sample we use data from the 2000 census, normal-
of neighbourhoods and in different regional ised to 2013 ZCTA boundaries using GIS,
contexts. and five-year estimates from the 2013
American Community Survey. As we note
above, we define gentrification as the process
Modelling the relationship by which neighbourhoods experiencing disin-
between the arts, gentrification vestment experience an influx of capital and
and displacement1 of middle and upper class residents. Although
specific definitions of gentrification vary in
Data
the literature, particular variables – income,
To analyse the relationships between arts education, new housing construction and
activity, arts growth, gentrification and housing value – are applied fairly consistently
neighbourhood displacement, we study zip in prior research to construct this definition
codes in the 30 US Metropolitan Statistical of gentrification. To model gentrification, we
Areas (2013 definitions) with a population follow Freeman’s (2005, 2009) established
over 2,000,000 between 2000 and 2013 approach and statistically distinguish neigh-
(Appendix, Table A3). We study the largest bourhoods with the potential to gentrify from
metros because on average they contain others based on a set of features that are com-
larger pools of arts employment and possess mon across definitions of gentrification in the
more neighbourhoods that experience gen- literature. Specifically, a zip code has poten-
trification compared to the majority of tial to gentrify if it has:
smaller metros. We choose zip codes for our
analysis because gentrification and displace-  a median income in 2000 that is less than
ment are both processes that occur at the the metro median in 2000; and
neighbourhood-level and the zip code is the  a percentage of new housing built in
smallest unit of analysis for which data on 1980 or later that is less than the metro
arts establishments is available at the area average in 2000.
national scale. Second, while zip codes
reflect postal routes rather than specific Next, from this pool of zip codes, we classify
units of geography, the census has developed those as gentrifying if they experience:

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

6 Urban Studies

 a growth (percent change) from 2000 to otherwise stable populations as a proxy for
2013 in the percent of the population displacement. While CCI notes that these
over 25 with a bachelor’s degree or conditions could instead indicate that lower
higher that is equal to or greater than income households are moving into higher
the metro area growth; and income categories, their analysis indicates a
 an increase in median housing value net increase in low income households over
from 2000 to 2013. the study period due to the Great Recession.
As a result, this approach may actually
We classify zip codes as gentrified if they underestimate displacement (Zuk, 2015).
experience the requirements for gentrifying Finally, following Grodach et al. (2014),
and have either: we create measures of two types of arts
establishments in 2000 – fine arts and com-
 a median housing value in 2013 that is mercial arts – and two measures of arts
greater than the metro average; or growth from 2000 to 2013. The two cate-
 a median household income in 2013 that gories of arts establishments are based on
is greater than the metro average. industry establishment counts from the North
American Industrial Classification System
We thus define three neighbourhood cate- (NAICS) in the 2000 and 2013 zip code busi-
gories, each of which corresponds to a dif- ness patterns (ZBP) dataset provided by the
ferent stage of gentrification.2 Importantly, US Census Bureau. We define fine arts estab-
a zip code cannot reach the stage of gentrify- lishments in 2000 as the sum of NAICS estab-
ing if it does not meet the requirements of lishment counts for theatre and performing
potential to gentrify. Similarly, a zip code arts companies, museums, art galleries, inde-
cannot be classified as gentrified if it does pendent artists and fine arts schools. We
not also meet the requirements of gentrifying define commercial arts establishments in 2000
and potential to gentrify. Alongside these as the sum of NAICS establishment counts
categories, in the descriptive analysis we also for film, music and design-related industries.
examine two types of neighbourhoods that Lastly, in addition to the number of establish-
do not have a potential to gentrify. Affluent ments in 2000, we also define growth in both
neighbourhoods are those in the top 20% of arts categories using the formula:
the metro’s median household income distri-  
bution. All remaining neighbourhoods that arts2013 + 1
ln ;
are not affluent and do not have a potential arts2000 + 1
to gentrify at the start of the study period
we classify as no potential to gentrify.3 Data where arts is the arts category, either fine
on neighbourhood characteristics by each of arts or commercial arts. This data set allows
the five neighbourhood categories are avail- us to study a range of establishments directly
able in the technical appendix (Tables A4 engaged in artistic production across many
and A5). metros at the neighbourhood level. However,
To define displacement, we identify zip although we include data on independent art-
codes that experienced a decrease in the num- ists, the NAICS is not an ideal measure since
ber of poor persons from 2000 to 2013. This only those who report as artists to the IRS
approach follows that of the University of are counted. That said, other work has found
California Berkeley’s Center for Community significant location correlations between com-
Innovation (CCI), which uses the loss of low mercial arts establishments, arts organisations
income households in neighbourhoods with and artists (Stern and Seifert, 2010).

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 7

Regression model however, does not account for all factors

Each of the above variables – gentrification, that may be associated with an arts presence,
displacement, fine arts and commercial arts gentrification and displacement, which may
in 2000 and fine arts and commercial arts confound significant results. To address this,
growth – is included in a path analysis shown we test the model with a set of control vari-
in Appendix Figure A1 and in Figure 8 in ables related to population, neighbourhood
the regression results section. As the diagram and built environment characteristics that
illustrates, the levels of fine and commercial prior work (e.g. Grodach et al., 2014;
arts in 2000 are hypothesised to predict gen- Woronkowicz, 2016) has found to be impor-
trification, displacement and future arts tant in arts-related neighbourhood change
growth. Gentrification is hypothesised to (Appendix, pp. 2–4).4
predict displacement and arts growth, and
displacement is hypothesised to predict arts
Region models
growth. Each pathway in the model has an
associated coefficient that indicates the rela- In addition to our primary research ques-
tionship between the independent and tions, we ask the following secondary ques-
dependent variables. We use maximum like- tion: Do the relationships between arts,
lihood estimation to obtain the model para- gentrification and displacement, and arts
meters and specify standard errors that are growth differ depending on region? To
robust to non-normal distributions and non- address this question, we run the above
independence of observations. The appendix model separately for four diverse regions:
elaborates on the equations used in the study New York-Newark-Jersey City, NY-NJ-PA;
to estimate the model, and includes more Chicago-Naperville-Elgin, IL-IN-WI; Dallas-
detailed model results. Fort Worth-Arlington, TX; and Los Angeles-
In our discussion of the results below we Long Beach-Anaheim, CA.
simplify the models’ interpretation by pre- These metropolitan areas are ideal study
senting the coefficients in terms of positives units because they possess different regional
and negatives, with positives indicating a and arts industry characteristics that allow
positive association between the independent more contextualised results, as well as a suf-
variable and the dependent variable, and ficient number of observations (zip codes)
negatives indicating the reverse. Full results for analysis. Los Angeles and New York are
are available in the Appendix, Table A8. established hubs of arts industry activity in
terms of employment size and concentration.
Chicago ranks third in the US in terms of
Control variables arts industry employment, but has only an
Our model captures the stage theory of gen- average employment concentration (LQ),
trification, which suggests that the arts move while Dallas contains a sizeable employment
to lower income neighbourhoods and in turn base but a very small concentration of
spur successive waves of gentrification, and employment (Grodach, 2016). New York
the theory that gentrification and displace- and particularly Los Angeles contain very
ment may alternatively spur growth in the high numbers of arts establishments per
arts. Since both of these processes are esti- neighbourhood while Chicago is roughly
mated simultaneously, our approach allows average and Dallas is well below the average
a more nuanced study of the relationship in the study population. Moreover, whereas
between arts, gentrification and displace- arts growth rates are very small or negative
ment than previously conducted. The model, in the latter metros, Los Angeles and New

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

8 Urban Studies

30

25
No potential to gentrify
20
Potential to gentrify
15 Gentrifying
Gentrified
10
Affluent
5

0
Fine Art Commercial Art

Figure 1. Average number of fine and commercial arts establishments by neighbourhood type, 2000.

York exceed average metro growth rates by and affluence is already present. The results
20% or more (see Appendix, Table A7). The also show that displacement is not limited to
model, however, does not account for all gentrifying areas, a phenomenon we explore
factors that may be associated with an arts in relation to the arts below.
presence, gentrification and displacement,
which may confound significant results. In
sum, Los Angeles and New York are estab- 30 metro statistics
lished arts regions with large numbers, high Contrary to the gentrification literature, fine
concentrations and large growth of the arts and commercial arts establishments are least
during the study period. Dallas represents common in gentrifying neighbourhoods
the other end of the spectrum and Chicago and those with the potential to gentrify
represents a more mixed arts industry con- (Figure 1). However, the commercial arts
text. Examining arts-gentrification patterns were sizably concentrated in areas that gen-
in these different contexts help us to under- trified over the study period yet these areas
stand nuances within the results based on account for just 12% of all study areas. This
the entire study population. could indicate that a commercial arts pres-
ence helps to encourage a process of rapid
upscaling. However, they also make a strong
Descriptive statistics: Art,
showing in affluent areas, indicating that
gentrification and displacement they may be attracted to and can afford
The descriptive statistics provide a portrait higher rent areas. This assumption is rein-
of the relationship between the arts, gentrifi- forced by the fine arts, which exhibit a simi-
cation and displacement that adds to the lar preference, but are most common in
complexity of the standard arts-led gentrifi- affluent areas. Still, both the fine and com-
cation narrative. Our analysis indicates that mercial arts are also common in areas we
fine and commercial arts industries tend to classify as having no potential to gentrify –
locate in more upscale areas and, if they do those that were well-off but not defined as
trigger gentrification and displacement, it is affluent at the start of the study period.
less likely to occur in predominately low About 24% of all areas in the study expe-
income areas than in those neighbourhoods rienced potential displacement. As we would
where gentrification is highly accelerated expect, signs of displacement are most

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 9

50 50

40 40

30 30
All neighborhoods All neighborhoods
20 20
Neighborhoods with Neighborhoods with
10 displacement 10 displacement

0 0

Figure 2. Average number of fine arts establishments (left) and commercial arts establishments (right) by
neighbourhood type, 2000.

160% 70%
140% 60%
120% 50%
100% 40%
80% 30%
60% 20%
40% All neighborhoods All neighborhoods
10%
20% 0%
0% Neighborhoods with -10% Neighborhoods with
displacement displacement

Figure 3. Average percent growth of fine arts establishments (left) and commercial arts establishments
(right) by neighbourhood type, 2000–2013.

common in gentrifying and gentrified areas, potential (Figure 3). Growth is strongest in
but those we label as affluent exhibit notable gentrified and affluent areas, and regardless
displacement as well (Appendix, Table A5). of displacement status. This indicates that
However, when we isolate the displacement the fine arts are attracted to these higher
neighbourhoods from the total sample, the rent markets or that initial arts establish-
arts remain weakest in areas suspected to be ments are replaced by more upscale arts
gentrifying (Figure 2). Also, an arts presence activities at a faster rate over the study
in higher income neighbourhoods with period. The commercial arts show basically
potential displacement stands out. In fact, the same pattern with one key difference –
the fine arts are nearly four times more con- they are on the decline in affluent neigh-
centrated in higher income places with signs bourhoods (Figure 3). Alongside the trends
of displacement than in those without. In above, this distinction is important because
other words, the highest concentrations of it potentially indicates that commercial arts
arts establishments are in places that are establishments are leaving affluent areas and
already wealthy and becoming increasingly contributing to the rapid upscaling of gentri-
segregated by wealth. fied places.
Examining the growth of arts establish-
ments reinforces this finding. The fine arts
exhibit stronger growth where signs of dis- Metro-specific trends
placement are present in all neighbourhood Trends in the four metropolitan areas gener-
types except those without gentrification ally align with the overall patterns, in that

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

10 Urban Studies

Figure 4. Average number of Chicago fine and commercial arts establishments by neighbourhood type.

Figure 5. Average number of Dallas fine and commercial arts establishments by neighbourhood type.

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 11

Figure 6. Average number of Los Angeles fine and commercial arts establishments by neighbourhood type.

arts establishments predominately locate turn, it could signal that the arts play a role
and grow in wealthier areas. Mirroring over- in the rapid upscaling of such places or that
all results, the arts are generally weakest in they move to seek out other areas where
the gentrifying areas and those with the incomes are growing.
potential to gentrify. There are some notable Los Angeles and New York (Figures 6
exceptions. These are perhaps a reflection of and 7) reveal contrasting dynamics com-
the high variation in the average number pared to the other metros. In Los Angeles,
and growth of arts establishments between the fine arts show by far the highest concen-
these metros. tration and growth in the metro’s nine afflu-
As in the full study, in Chicago and ent areas. The commercial arts are strong
Dallas, the arts are least common in areas but stagnant there and, by the end of the
that are gentrifying and that possess the study period, gentrified areas contain the
potential to gentrify (Figures 4 and 5). Yet, highest concentrations. New York stands
we also find that areas which gentrified dur- out because the highest arts concentrations
ing the study period generally held and are in areas with no potential to gentrify
gained the largest numbers of arts establish- (though there is noticeable growth in gentri-
ments. This is especially notable in Dallas fied areas as well). In other words, arts
where just five areas gentrified. In contrast industries are more dispersed within New
to the overall results, however, arts estab- York rather than concentrated predomi-
lishments are far less common in affluent nately in gentrified or affluent areas.
areas. This may indicate that cost pressures Figures 4–7 also highlight that there is
in affluent areas are a barrier and push them not a significant difference in the location or
towards those that gentrified since 2000. In growth of arts industries based on

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

12 Urban Studies

Figure 7. Average number of New York fine and commercial arts establishments by neighbourhood type.

neighbourhood displacement status. In Los Regression results: Are arts

Angeles and New York, arts industries are industries linked with
more numerous and growing slightly faster gentrification and displacement?
in areas with potential displacement. While
Dallas’ commercial arts establishments Overall, the regression models indicate that
favoured displacement areas, they are on the arts industries generally do not play a signifi-
decline there. However, the region also expe- cant role in gentrification and displacement.
rienced an overall decline in arts establish- The presence of commercial arts establish-
ments during the study period. ments does increase the likelihood of gentrifi-
In summary, the descriptive statistics cation, but only weakly. A fine arts presence,
show that arts industries are largely located on the other hand, has a negative yet insignif-
and growing fastest in gentrified and higher icant relationship to gentrification. Neither
income places. In most metros, they also fine arts nor commercial arts appear to have
tend to have higher numbers in places where a direct relationship with displacement.
displacement occurs. However, based on Moreover, the regression analyses show that
these results, we cannot say whether the arts contextual differences and type of art matter
drive gentrification and displacement or are in relation to these neighbourhood pressures.
attracted to more upscale areas where dis- Figure 8 shows the direct relationships
placement might occur. In the next section, between the arts, gentrification and displace-
we control for neighbourhood conditions to ment based on the path analysis. Table 1
determine how the arts may impact gentrifi- shows the basic results for the models used
cation and displacement. to estimate the total effects (direct and

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 13

Figure 8. The relationship between arts, gentrification and displacement: Direct effects (full results in
Appendix, Table A7).

Table 1. The relationship between arts, gentrification and displacement: Significance and direction of total
effects (full results in Appendix, Table A8).

All metros Chicago Dallas Los Angeles New York

 Gentrification
1. Fine arts establishments + +
2. Commercial arts establishments + +
 Displacement
3. Fine arts establishments +
4. Commercial arts establishments –
5. Gentrification + +
 Fine arts growth
6. Gentrification + + + + +
7. Displacement + +
 Commercial arts growth
8. Gentrification + + + + +
9. Displacement
Covariance of arts growth variables + + +
N 996 46 52 77 106
R2finegrow 0.14 0.24 0.22 0.13 0.18
R2comgrow 0.13 0.14 0.29 0.07 0.11
Pseudo R2disp 0.15 0.19 0.21 0.06 0.03
Pseudo R2gent 0.08 0.29 0.15 0.05 0.01

Notes: + significant positive coefficient; – significant negative coefficient; blank = insignificant coefficient. Estimate is
considered significant if p \ 0.1.

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

14 Urban Studies

indirect) of these relationships for neigh- Chicago and Dallas respectively). While we
bourhoods in each of the four regions com- would expect the process to be more wide-
pared to neighbourhoods in all 30 regions. spread in Los Angeles and New York given
The models indicate that a fine arts presence their sizable arts concentrations and competi-
does not increase the likelihood of either tive real estate markets, we do not find a
gentrification or displacement in the 30 significant relationship. Instead, the arts-
metros overall (Table 1, rows 1 and 3 respec- gentrification link is strongest, though by no
tively). In contrast, an increase in the pres- means primary, when the arts are specifically
ence of commercial arts establishments concentrated in a handful of gentrified areas,
increases the likelihood of gentrification by as the descriptive statistics reveal for Chicago
0.29. However, the relationship between the and especially Dallas. In contrast, in Los
commercial arts and displacement is insignif- Angeles and New York where the arts are
icant. These findings are in line with prior more dispersed throughout the different
research (Grodach et al., 2014). However, types of neighbourhoods, gentrification is
the results suggest that although the com- driven predominately by other factors.
mercial arts may catalyse gentrification, dis- The individual metro results tell little in
placement is not inextricably linked to the terms of displacement. All results are insig-
presence of the arts. Rather, it is important nificant with the exception of those in New
to recognise that the arts are just one of many York. Here, in contrast to the overall
other complex factors that account for gentrifi- results, the fine arts have a significant effect
cation and that any subsequent displacement is on displacement and the commercial arts
likely due to other dimensions. In fact, given have a negative association. The fact that
the very low R2 (0.08) for gentrification in the these results exactly oppose that of the 30
full model, we are likely only accounting for a metros overall may be due to New York’s
small slice of factors that explain the gentrifica- highly concentrated and diverse arts sector
tion process. The model does do a better job at that sets it apart from most places. Still, the
predicting where displacement may occur (R2 low R2 for gentrification (0.01) and displace-
= 0.15). ment (0.03) in New York suggests that the
The individual metro results reveal impor- arts are very weak predictors of these forms
tant contextual differences in the relationship of neighbourhood change. The contrast with
of the arts to gentrification and displace- the classic case studies of artist-induced gen-
ment. Interestingly, it is in the metros that trification and displacement, many of which
have fewer arts establishments and a lower were conducted in New York, is significant
proportion of neighbourhoods experiencing (c.f. Deutsche and Ryan, 1984; Zukin,
gentrification where the link between gentri- 1982). Considering New York City’s contin-
fication and the arts is present. The fine arts ued investment in art facilities such as the
have a significant effect on gentrification in construction of the new Whitney Museum
Chicago and Dallas but not in Los Angeles and the Brooklyn Cultural District, it is pos-
and New York (Table 1, row 1). The com- sible that this relationship between the fine
mercial arts have a significant effect on gen- arts and displacement may be partly driven
trification only in Chicago. Moreover, the by explicit economic development strategies,
models explain considerably more of the var- as opposed to the presence of the arts
iation in gentrification and displacement in themselves.
these two metros than they do in the major Examining the relationships from the
art hubs (the R2 is 0.29 and 0.15 for gentrifi- other direction, the model indicates that in
cation and 0.19 and 0.21 for displacement for the 30 metros, gentrification predicts the

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 15

growth of fine and commercial arts estab- near future, pricing out lower income resi-
lishments (Table 1, rows 6 and 8). In other dents and businesses in these individual
words, while arts growth does occur in the places. In contrast, given that New York has
context of gentrification, an arts presence is more neighbourhoods undergoing gentrifica-
not driving the relationship. Rather, gentri- tion and more arts establishments prefer gen-
fied environments lead to arts growth. This trified spaces, the process is likely more
holds for results across the four metros, with widespread, as a large literature has demon-
the exception of New York fine arts where strated. But, in contrast to the standard nar-
the relationship is insignificant. rative, given our results, it is more likely that
Neighbourhoods with potential displace- arts establishments are attracted to rather
ment similarly predict fine arts establishment than cause gentrification here.
growth in the 30-metro sample, yet in our
four metros this phenomenon is only signifi- Conclusion
cant for the fine arts in New York. As
growth in fine arts appears to have been This study contributes to a deeper under-
concentrated in higher income areas, these standing of the relationship between the arts,
findings suggest that arts establishments are gentrification, and displacement. Chiefly, we
locating in already upscale neighbourhoods provide evidence that the standard arts-
that are only becoming wealthier. Further, gentrification narrative is not a generalised
this may be due to explicit arts-oriented pattern, but rather one possible way that the
development projects as discussed above. arts are implicated in place change. By going
Yet, again, the R2 for displacement is quite beyond the focus on the role of artists in
low. As such, results indicate that although gentrification, this study adds a new dimen-
the arts are attracted to places that are sion to understanding how the arts affect
upgrading, the association with displace- place change and provides an important cor-
ment is not clear-cut. rective to the stage model of gentrification.
Finally, the metro analysis reveals diver- The analysis demonstrates that fine and
gent relationships between gentrification commercial arts establishments have varied,
and displacement that are worth considering conflicting relationships with gentrification
compared to the full study. Namely, gentrifi- and displacement in different places. In par-
cation has a significant effect on displacement ticular, arts-led urban policy (the apotheosis
only in Dallas (Table 1, row 5). However, as of the stage model) breaks down because, as
noted above, of the four metros under study, our results demonstrate, arts establishments
Dallas has few areas where these neighbour- by and large do not predict gentrification.
hood change processes occur – five neigh- Given that the arts have become consider-
bourhoods gentrified and about 10 (out of 92 ably more diverse in their economic and
neighbourhoods) experienced potential displa- social impacts and their participants, our
cement during the study period. In other findings provide a significant contribution
words, although gentrification and displace- to understanding these wider dynamics.
ment are linked in Dallas, the process is con- Overall, gentrifying areas actually have
fined to a few areas. As we note above, an the weakest arts presence of all neighbour-
overarching arts-displacement relationship is hood types in the study. While commercial
insignificant in Dallas. However, because gen- arts establishments have a strong share in
trification is confined to a few neighbour- those areas that gentrified over the study
hoods, the arts may have more of an impact period, they are equally represented in
on bidding up real estate values and, in the places that show no potential to gentrify.

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

16 Urban Studies

The fine arts are most numerous in affluent regional arts activity in combination with a
areas and those without potential to gen- competitive real estate market that explains
trify. Both the fine arts and commercial arts-related gentrification.
arts do exhibit the highest growth rates in These findings provide an important les-
gentrifying and gentrified places and these son to policymakers who seek to use the arts
rates are highest where there are signs of and creative industries in place-based revita-
displacement. Yet, our path analysis shows lisation programs. Given that arts establish-
that it is more likely that certain arts estab- ments cluster amongst linked industries and
lishments are attracted to these environ- their customer base, it is folly to assume pol-
ments rather than creating them. In short, icy can seed arts establishments and produce
the arts likely do not play a causal role in place improvements. Nor, however, is it
likely that arts-led strategies will engender
gentrification and displacement.
gentrification and displacement. The caveat
Rather, the typical process is often turned
to this is that regions without highly concen-
on its head – instead of sparking gentrifica-
trated arts activity are more likely to experi-
tion, arts establishments are most numerous
ence gentrification through the arts in those
in places that are well-off and upscaling.
few places with highly concentrated arts
One could argue that this is because they
industries. In such instances in particular,
have already contributed to gentrifying their
policy should attend to potential displace-
locations. However, if they play a role in the
ment pressures.
standard arts-gentrification narrative, then
we should also see the movement of this arts
activity to lower rent areas where gentrifica- Declaration of conflicting interests
tion and displacement continue anew. We The author(s) declared no potential conflicts of
largely do not. Arts growth is very small in interest with respect to the research, authorship,
places that are gentrifying or exhibit the and/or publication of this article.
potential to gentrify, particularly for the
commercial arts. This indicates that many Funding
arts establishments want to be near their cli-
ents, whether it is the patron base of a gal- Funding for this research was generously pro-
vided by the National Endowment for the Arts
lery, or a design firm that serves a
(Grant 12-3800-7004).
technology or advertising company, rather
than simply seeking out cheap rent.
These patterns largely hold in the four Notes
metros we study individually, but with an 1. Due to space restrictions, a technical appen-
interesting deviation. It is the concentration of dix containing a detailed description of all
arts activity in particular neighbourhoods in data, descriptive statistics related to model
regions where the arts and gentrification are variables and neighbourhood characteristics,
less widespread that explains arts-led gentrifi- as well as regression methodologies is avail-
cation. As we show, an arts industry presence able at: http://eprints.qut.edu.au/101715/.
2. Researchers replicating our definitions of gen-
predicts gentrification in Chicago and Dallas,
trification should be careful if their sample
but not in Los Angeles and New York as we
includes a large number of rural zip codes
would expect. We surmise this is because of (our sample does not). The definitions we use
the concentrated nature of gentrification out- could classify developing rural areas as gentri-
side all but the largest, most expensive metros, fying even though they would not have suf-
thus magnifying an arts-gentrification effect. fered from the disinvestment that provides
As such, it is not the size or concentration of the context for gentrification.

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 17

3. Although Freeman’s approach distinguishes city’ in mixed-income communities: Uses and

between neighbourhoods with potential to expectations of space and place. International
gentrify and neighbourhoods experiencing Journal of Urban and Regional Research 37:
gentrification, it does not distinguish between 480–502.
neighbourhoods that have never experienced Curran W (2010) In defense of old industrial
gentrification and those in which gentrifica- spaces: Manufacturing, creativity and innova-
tion occurred at a period prior to the study tion in Williamsburg, Brooklyn. International
period. Affluent neighbourhoods may have Journal of Urban and Regional Research 34(4):
experienced gentrification at some point prior 871–885.
to the study period and increases in affluence Deutsche R and Ryan CG (1984) The fine art of
in the study period could thus be extensions gentrification. October 31: 91–111.
of that process. Despite these limitations, the Foster N, Grodach C and Murdoch J (2016)
definitions clearly distinguish between neigh- Neighborhood diversity, economic health and
bourhoods that are experiencing gentrifica- the role of the arts. Journal of Urban Affairs.
tion during the study period and those that Epub ahead of print, 25 January 2016. DOI:
are not. This allows for a statistically rigorous 10.1111/juaf.12278.
examination of the relationship between the Freeman L (2005) Displacement or succession?
arts, gentrification and displacement. Residential mobility in gentrifying neighbor-
4. The presence of arts or gentrification in sur- hoods. Urban Affairs Review 40(4): 463–491.
rounding zip codes is an important factor that Freeman L (2009) Neighborhood diversity, met-
we do not consider in this study. Future work ropolitan segregation and gentrification: What
may wish to investigate this issue. are the links in the US? Urban Studies 46(10):
2079–2101.
References Freeman L and Braconi F (2004) Gentrification
and displacement: New York City in the
Ashley AJ (2014) Negotiating risk in property- 1990s. Journal of the American Planning Asso-
based arts economic development: Exploring ciation 70(1): 39–52.
the innovative but untimely development part- Glass R (1964) Introduction: Aspects of change.
nership between the Seattle Art Museum and In: The Centre for Urban Studies (ed.) Lon-
Washington Mutual. Cities 37: 92–103. don: Aspects of Change. London: MacGibbon
Bridge G, Butler T and Lees L (eds) (2011) Mixed & Kee, pp. xiii–xlii.
Communities: Gentrification by Stealth? Bris- Grodach C (2010) Beyond Bilbao: Rethinking
tol: Policy Press. flagship cultural development and planning in
Brown-Saracino J (2010) A Neighborhood that three California cities. Journal of Planning
Never Changes: Gentrification, Social Preser- Education and Research 29(3): 353–366.
vation, and the Search for Authenticity. Chi- Grodach C (2011) Art spaces in community and
cago, IL: University of Chicago Press. economic development: Connections to neigh-
Cameron S and Coaffee J (2005) Art, gentrifica- borhoods, artists, and the cultural economy.
tion, and regeneration – From artist as pioneer Journal of Planning Education and Research
to public arts. European Journal of Housing 31(1): 74–85.
Policy 5(1): 39–58. Grodach C (2012) Before and after the creative city:
Catungal JP, Leslie D and Hii Y (2009) Geogra- The politics of urban cultural policy in Austin,
phies of displacement in the creative city: The Texas. Journal of Urban Affairs 34(1): 81–97.
case of Liberty Village, Toronto. Urban Stud- Grodach C (2013) Cultural economy planning in
ies 46(5–6): 1095–1114. creative cities: Discourse and practice. Interna-
Chang TC (2016) ‘‘New uses need old buildings’’: tional Journal of Urban and Regional Research
Gentrification aesthetics and the arts in Singa- 37(5): 1747–1765.
pore. Urban Studies 53(3): 524–539. Grodach C (2016) Mapping the arts: Industry
Chaskin RJ and Joseph ML (2013) ‘Positive’ gen- concentrations, distribution, and change in the
trification, social control and the ‘right to the

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

18 Urban Studies

US, 1980–2010. International Journal of Cul- crime rates in Chicago, 1991 to 2005. City and
tural Policy 22(3): 353–375. Community 10(3): 215–240.
Grodach C, Foster N and Murdoch J (2014) Gen- Ponzini D and Rossi U (2010) Becoming a crea-
trification and the artistic dividend: The role of tive city: The entrepreneurial mayor, network
the arts in neighborhood change. Journal of the politics and the promise of an urban renais-
American Planning Association 80(1): 21–35. sance. Urban Studies 47(5): 1037–1057.
Hackworth J and Smith N (2001) The changing Pratt A (2009) Urban regeneration: From the arts
state of gentrification. Tijdschrift voor Econo- ‘feel good’ factor to the cultural economy: A
mische en Sociale Geografie 94(4): 464–477. case study of Hoxton, London. Urban Studies
Hutton TA (2009) The New Economy of the Inner 46(5–6): 1041–1061.
City: Restructuring, Regeneration and Disloca- Sasajima S (2013) From red light district to art
tion in the 21st Century Metropolis. Abingdon: district: Creative city projects in Yokohama’s
Routledge. Kogane-cho neighborhood. Cities 33: 77–85.
Lees L (2003) Super-gentrification: The case of Schuetz J (2014) Do art galleries stimulate redeve-
Brooklyn Heights, New York City. Urban lopment? Journal of Urban Economics 83:
Studies 40(12): 2487–2509. 59–72.
Lees L, Slater T and Wyly E (2008) Gentrification. Schuetz J and Green R (2014) Is the art market
London: Routledge. more bourgeois than bohemian? Journal of
Ley D (1996) The New Middle Class and the Regional Science 54(2): 273–303.
Remaking of the Central City. Oxford: Oxford Shaw K (2008) Gentrification: What it is, why it
University Press. is, and what can be done about it. Geography
Ley D (2003) Artists, aestheticisation and the Compass 2(5): 1697–1728.
field of gentrification. Urban Studies 40(12): Shaw K (2013) Independent creative subcultures
2527–2544. and why they matter. International Journal of
Lloyd R (2010) Neo-Bohemia: Art and Commerce Cultural Policy 19(3): 333–352.
in the Postindustrial City. 2nd edn. New York: Silver D and Miller D (2013) Contextualizing the
Routledge. artistic dividend. Journal of Urban Affairs
Markusen A and Gadwa A (2010) Arts and cul- 35(5): 591–606.
ture in urban or regional planning: A review Slater T (2006) The eviction of critical perspec-
and research agenda. Journal of Planning Edu- tives from gentrification research. International
cation and Research 29(3): 379–391. Journal of Urban and Regional Research 30(4):
Mathews V (2010) Aestheticizing space: Art, gen- 737–757.
trification and the city. Geography Compass Smith N (1996) The New Urban Frontier: Gentri-
4(6): 660–675. fication and the Revanchist City. London:
Mathews V (2014) Incoherence and tension in Routledge.
culture-led redevelopment. International Journal Stern M and Seifert S (2010) Cultural clusters:
of Urban and Regional Research 38: 1019–1036. The implications of cultural assets agglomera-
Murdoch J, Grodach C and Foster N (2016) The tion for neighborhood revitalization. Journal
importance of neighborhood context in of Planning Education and Research 29(3):
arts-led development: Community anchor or 262–279.
creative class magnet? Journal of Planning Woronkowicz J (2016) Art-making or place-mak-
Education and Research 36(1): 32–48. ing? The relationship between open-air perfor-
Newman K and Wyly EK (2006) The right to stay mance venues and neighborhood change.
put, revisited: Gentrification and resistance to Journal of Planning Education and Research
displacement in New York City. Urban Studies 36(1): 49–59.
43(1): 23–57. Wyly EK and Hammel DJ (1999) Islands of decay
Papachristos A, Smith C, Scherer M, et al. (2011) in seas of renewal: Housing policy and the
More coffee, less crime? The relationship resurgence of gentrification. Journal of Hous-
between gentrification and neighborhood ing Policy 10(4): 711–771.

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016

Grodach et al. 19

Yoon H and Currid-Halkett E (2015) Industrial Zukin S (1982) Loft Living: Culture and Capital
gentrification in West Chelsea New York: in Urban Change. Baltimore, MD: Johns Hop-
Who survived and who did not? Empirical evi- kins University Press.
dence from discrete-time survival analysis. Zukin S (2010) Naked City: The Death and Life of
Urban Studies 52(1): 20–49. Authentic Urban Places. Oxford: Oxford Uni-
Zuk M (2015) Regional early warning system for versity Press.
displacement typologies: Final project report. Zukin S and Braslow L (2011) The life cycle of
Report, Center for Community Innovation, New York’s creative districts: Reflections on
University of California Berkeley, USA, 23 the unanticipated consequences of unplanned
July. Available at: http://iurd.berkeley.edu/ cultural zones. City, Culture and Society 2(3):
uploads/CCI_Final_Report_07_23_15.pdf 31–140.
(accessed 8 March 2016).

Downloaded from usj.sagepub.com at Queensland University of Tech on December 7, 2016