Access to this full-text is provided by Springer Nature.
Content available from BMC Public Health
This content is subject to copyright. Terms and conditions apply.
R E S E A R C H Open Access
Who engages in the arts in the United
States? A comparison of several types of
engagement using data from The General
Social Survey
Jessica K. Bone
1*
, Feifei Bu
1
, Meg E. Fluharty
1
, Elise Paul
1
, Jill K. Sonke
2
and Daisy Fancourt
1
Abstract
Background: Engaging in the arts is a health-related behavior that may be influenced by social inequalities. While
it is generally accepted that there is a social gradient in traditional arts and cultural activities, such as attending
classical music performances and museums, previous studies of arts engagement in the US have not adequately
investigated whether similar demographic and socioeconomic factors are related to other forms of arts
engagement.
Methods: Using cross-sectional data from the General Social Survey (GSS) in the US, we examined which
demographic, socioeconomic, residential, and health factors were associated with attendance at arts events,
participation in arts activities, membership of creative groups, and being interested in (but not attending) arts
events. We combined data from 1993 to 2016 in four analytical samples with a sample size of 8684 for arts events,
4372 for arts activities, 4268 for creative groups, and 2061 for interested non-attendees. Data were analysed using
logistic regression.
Results: More education was associated with increased levels of all types of arts engagement. Parental education
demonstrated a similar association. Being female, compared to male, was also consistently associated with higher
levels of engagement. Attendance at arts events was lower in participants with lower income and social class,
poorer health, and those living in less urban areas. However, these factors were not associated with participation in
arts activities or creative groups or being an interested non-attendee.
Conclusions: Overall, we found evidence for a social gradient in attendance at arts events, which was not as
pronounced in participation in arts activities or creative groups or interest in arts events. Given the many benefits of
engagement in the arts for education, health, and wider welfare, our findings demonstrate the importance of
identifying factors to reduce barriers to participation in the arts across all groups in society.
Keywords: Arts, Culture, Social gradient, Wellbeing, Health, United States
© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data.
* Correspondence: jessica.bone@ucl.ac.uk
1
Research Department of Behavioural Science and Health, Institute of
Epidemiology & Health, University College London, London, UK
Full list of author information is available at the end of the article
Bone et al. BMC Public Health (2021) 21:1349
https://doi.org/10.1186/s12889-021-11263-0
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Background
There are many known social inequalities in health, in-
cluding differences in healthy life expectancy and mor-
tality [1,2]. These disparities may be partially explained
by a social gradient in a variety of health behaviors, in-
cluding diet, obesity, physical activity, alcohol consump-
tion, and smoking [3–5]. Health behavior norms may be
learnt within the socioeconomic context, with social de-
terminants influencing behavior throughout the life
course [6]. Engaging in the arts is an example of a
health-related behavior that demonstrates social inequal-
ities [7,8].
Arts engagement typically refers to different types of
creative activity, from actively participating in the arts
(e.g. dancing, singing, acting, painting, reading) to more
receptive cultural engagement (e.g. going to museums,
galleries, exhibits, performances and the theater [9]). It
can also encompass broader creative activities that,
whilst not always labelled as ‘arts’, share similar proper-
ties of creative skill and imagination (e.g. gardening,
cooking, and hobby or book groups [10]). In 2019, the
World Health Organization identified more than 3000
studies showing the beneficial impact of arts engagement
on mental and physical health and social determinants
of health, from education to social cohesion and welfare
[9]. Despite growing awareness of the benefits of en-
gaging with the arts, there is a social gradient in arts par-
ticipation. Previous surveys have found that arts
engagement in the United States (US) may differ accord-
ing to socioeconomic status, education, and income
[11–13]. Similar factors are associated with inequalities
in access to health care and health and social outcomes
[14–17]. Varying engagement in the arts may therefore
further contribute to health and social inequalities [8].
However, the literature on this topic is limited by a
number of factors.
First, many previous studies have focused on certain
demographic or socioeconomic predictors of arts en-
gagement without always taking into account the broad
range of factors that may be related to one another.
From these studies, the most consistent predictors of in-
creased arts engagement are higher levels of education
and income [12,13,18–24]. There have been extensive
efforts to differentiate the effects of education and in-
come on arts engagement, and it appears that both inde-
pendently contribute to engagement levels [21,25].
However, education may be more strongly associated
with attending highbrow cultural events, whereas in-
come is more strongly associated with other forms of
arts engagement [25]. Further, self-identified social class
may be another important factor which should be stud-
ied alongside income and education [23]. There is also
evidence for lower rates of engagement in Black than
White racial/ethnic groups [12,18,22,26,27]. Still, it
remains unclear whether race/ethnicity has a strong as-
sociation with engagement after other factors, particu-
larly education and income (as interconnected systems
that contribute to structural racism), have been taken
into account [18,21,22,27,28].
Additionally, there are other factors that could be as-
sociated with arts engagement that have not been inves-
tigated in the US to date. In the UK, there are
geographical differences in participation independent of
individual demographic and socio-economic back-
grounds [29]. Further, living alone is associated with
fewer perceived opportunities to engage in the arts and
those with poorer physical and mental health may ex-
perience more barriers to engaging [30]. As many previ-
ous studies of arts engagement in the US are based on
the Survey of Public Participation in the Arts (SPPA; Na-
tional Endowment for the Arts), which does not collect
data on physical and mental health, these factors have
not been investigated.
Moreover, in the US, most research on predictors of
arts engagement has measured engagement with ‘bench-
mark’arts activities, as defined in the SPPA. These activ-
ities include attending jazz, classical music, opera,
musical or non-musical plays, ballet performances, and
art museums or art galleries. Although these activities
are not intended to be comprehensive [31], they have re-
peatedly been used as a metric of engagement in the
arts. This has led to the perception that arts participa-
tion is declining in the US [11,22,32]. However, when
defined more broadly, including other types of arts activ-
ities and going beyond the non-profit sector to recognize
the many diverse commercial forms of cultural expres-
sion, participation is not declining and the way in which
people participate may instead be changing [13,33,34].
There may be a growing gap between arts participation
metrics and the ways in which people participate, and
this could be affecting our understanding of the predic-
tors of engagement [35].
Therefore, in this study, we used a large nationally
representative sample of adults in the US (the General
Social Survey; GSS) to investigate predictors of different
types of arts engagement. Specifically, we were interested
in whether there are social inequalities in engagement in
the arts, as found in other health-related behaviors. To
do this, we tested which demographic, socioeconomic,
residential, and health factors were associated with at-
tendance at arts events, participation in arts activities,
and membership of creative groups. Further, in order to
differentiate between non-attendance due to a lack of
interest versus non-attendance due to barriers or a lack
of opportunities, we investigated whether similar factors
were associated with being interested in, but not attend-
ing, arts events. Finally, we examined whether engage-
ment changed across time, from 1993 to 2016, and
Bone et al. BMC Public Health (2021) 21:1349 Page 2 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
whether associations between demographic and socio-
economic factors and engagement changed over these
two decades.
Methods
Sample
Participants were drawn from the General Social Survey
(GSS); a repeated cross-sectional and rotating panel
study of adults aged 18 and over in the US [36]. Each
survey year was an independently drawn sample of
English-speaking individuals living in non-institutional
arrangements. From 2006 onwards, Spanish-speakers
were added to the target population. Full probability
sampling was employed, and surveys sub-sampled non-
respondents from 2004 onwards.
We used data from GSS waves at which arts outcomes
were measured between 1993 and 2016. Each wave in-
cluded a unique sample of individuals so we were able
to combine data across waves. We used four indicators
of arts engagement (arts events, arts activities, creative
groups, and interested non-attendees), each measured in
different waves of the GSS. Arts events were measured
in 1993, 1998, 2002, 2012 and 2016, arts activities were
measured in 1993, 1998, and 2002, creative groups were
measured in 1993, 1994, 2004, and 2010, and interested
non-attendees were measured in 2012 and 2016. We
therefore identified four samples, one for each outcome.
When combining samples across all relevant years, the
total number of participants was 14,890, 7203, 12,311,
and 7687 for arts events, activities, creative groups, and
interested non-attendees respectively. We then restricted
the sample just to participants with complete data on
arts variables, which produced a final sample size of
8684 for arts events, 4372 for arts activities, 4268 for
creative groups, and 2061 for interested non-attendees
(see Supplementary Table 1for further details).
All participants gave informed consent and this study
has Institutional Review Board approval from the Uni-
versity of Florida (IRB201901792) and ethical approval
from University College London Research Ethics Com-
mittee (project 18839/001).
Arts engagement outcomes
Arts events
Participants were asked whether they had attended arts
events in the last 12 months, not including school per-
formances. In 1993, attendance at three events was mea-
sured as the following: a) art museum or gallery, b)
ballet or dance performance, and c) classical music or
opera performance. In 1998 and 2002, two additional
events were added to this list: d) popular music perform-
ance, and e) non-musical stage play performance. In
2012 and 2016, attendance at two types of event was
measured; a) music, theatre, or dance performance, and
b) art exhibit (including paintings, sculpture, textiles,
graphic design, or photography). Due to these differ-
ences in measurement across years, we collapsed all re-
sponses into a binary variable indicating attendance at
any event in the last 12 months (0 = none, 1 = one or
more). As this does not entirely account for the changes
in question style, we tested whether the changing defin-
ition of arts events altered our findings in sensitivity ana-
lyses (outlined below). For full details of the questions
asked in each wave, see Supplementary Table 2.
Arts activities
Participants self-reported whether they participated in
any kind of arts activity in the last 12 months, including:
a) making art or craft objects, b) taking part in music,
dance, or theatrical performance, and c) playing a mu-
sical instrument (Supplementary Table 2). This was
coded as a binary variable (0 = none, 1 = one or more),
and was measured consistently in 1993, 1998, and 2002.
Creative groups
Participants were asked about the groups or organiza-
tions of which they were a member in 1993, 1994, 2004,
and 2010. The creative groups were hobby or garden
clubs and literary, art, discussion, or study groups (Sup-
plementary Table 2). A binary variable was created indi-
cating membership in either of these group types (0 =
none, 1 = one or more).
Interested non-attendees
In the 2012 and 2016 GSS, participants who responded
to the arts event questions were also asked if there was
an arts event during the last 12 months that they had
wanted to go to but did not attend (0 = no, 1 = yes). In
2012, only participants who had not attended an event
during the last 12 months were asked this question. In
2016, all participants who were asked about arts event
attendance were also asked whether there was an event
that they had wanted to go to but did not attend. As we
aimed to include only participants who were interested
non-attendees, we excluded those who reported attend-
ing an arts event in 2016 (n= 738 excluded).
Exposures
We examined whether a range of demographic, socio-
economic, residential, and health factors were associated
with arts engagement. Demographics included age
(years), sex (male or female), race/ethnicity (White,
Black, or Other) and marital status (married, separated/
divorced/widowed, or never married). Socioeconomic
factors included total number of years of education (0–
20 years), parental years of education (highest reported
maternal or paternal education; 0–20 years), employ-
ment status in the last week (employed, unemployed or
Bone et al. BMC Public Health (2021) 21:1349 Page 3 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
not currently working, retired, keeping house, or other),
family income in constant dollars (base = 1986; $0 to
$9999, $10,000 to $19,999, $20,000 to $29,999, $30,000
to $49,999, or $50,000+), subjective satisfaction with fi-
nancial situation (not satisfied at all, more or less satis-
fied, or pretty well satisfied), and a subjective rating of
social class (lower class, working class, middle class, or
upper class).
Residential factors included level of urbanicity
(medium to large city with 50,000 people or more; sub-
urb of a medium to large city; unincorporated area of a
medium to large city; small city, town or village of 2500
to 50,000 people; and smaller areas or open country),
number of people living in the household (1–10), and
whether there was an area within a mile of their home
where they would be afraid to walk alone at night (yes
or no).
Finally, we included a general health rating (excellent,
good, fair, or poor).
Statistical analyses
We used four logistic regression models to test cross-
sectional associations between demographic, socioeco-
nomic, residential, and health exposures and binary arts
engagement outcomes. Each outcome (arts events, arts
activities, creative groups, interested non-attendees) was
modelled separately. Where there was evidence of a
non-linear association between age and arts engagement,
we included a quadratic age term. As a number of simi-
lar exposures were included, multicollinearity was
assessed to ensure that Variance Inflation Factors were
less than 10 [37]. All analyses were weighted to account
for the sub-sampling of non-respondents and the num-
ber of adults in the household using weights supplied by
the GSS [36]. We accounted for clustering of partici-
pants within primary sampling units by using robust
standard errors.
We also tested whether there was any evidence that
associations between arts engagement outcomes and
age, race/ethnicity, class, income, and sex differed over
time. We included an interaction term between each ex-
posure and survey year in separate logistic regression
models. Where there was evidence for an interaction, we
then examined the association between the exposure
and arts engagement separately in each survey year.
For participants with missing data on exposures, we
imputed data using multiple imputation by chained
equations (MICE [38]). We used linear, logistic, ordinal,
and multinomial regression and predictive mean match-
ing according to variable type, generating 50 imputed
data sets (maximum missing data ranged from 10 to
35% in each sample; Supplementary Table 3). The im-
putation model included all variables used in analyses,
auxiliary variables, and the survey weights. Auxiliary
variables were split ballot group, interviewer’s rating of
the respondent’s attitude toward the interview and un-
derstanding of questions, respondent’s rating of their
family income (relative to other Americans), and geo-
graphic mobility since age 16. Imputations were per-
formed separately according to survey year. For creative
groups, several exposures (satisfaction with financial
situation, general health rating, and feeling afraid in
neighborhood) and an auxiliary variable (relative in-
come) were missing for all participants in some years so
were not included in the imputations or analyses. All
other variables were successfully imputed. The results of
analyses did not vary between complete cases and im-
puted data sets (Supplementary Table 4), so findings
from the imputed data are reported. All analyses were
performed using Stata 16 [39].
Sensitivity analysis
We tested whether the changing definition of arts event
attendance altered our findings. In this analysis, we used
the most homogenous measures of arts events, those in-
cluded from 1998 to 2016. We therefore repeated the
main analysis excluding participants from 1993 (which
used a narrower definition of arts events) and examined
whether similar factors were associated with arts event
attendance in this subsample (n= 7094; Supplementary
Table 7).
Results
Arts events
In total, 8684 participants provided data on attendance
at arts events, 53% of whom were female and 78% were
White (Table 1). These participants ranged in age from
18 to 89 years, with a mean age of 46.6 (SD = 17.0).
Overall, 56% had attended an arts event in the last 12
months, although this varied across years (1993: 48%,
1998: 62%, 2002: 66%, 2012: 46%, 2016: 50%).
In the logistic regression model, there was evidence
for associations between several demographic factors
and attending arts events (Table 2). Females had 24%
higher odds (95% CI = 1.10–1.39) of attendance than
males. In comparison to White participants, Black par-
ticipants had 34% lower odds (95% CI = 0.55–0.78) of at-
tendance. Participants who had never been married had
37% higher odds (95% CI = 1.14–1.63) of attendance
than those who were married.
There was evidence that several socioeconomic factors
were associated with attendance. Compared to those
with a family income of less than $10,000, participants
in all other income groups had higher odds of attend-
ance. The highest odds were in the highest income
group. Subjective rating of social class was also associ-
ated with attendance, with higher classes associated with
increasing odds. Each additional year of education was
Bone et al. BMC Public Health (2021) 21:1349 Page 4 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Table 1 Demographic characteristics of the samples, with data combined across all included survey years
Events
n= 8684
Activities
n= 4372
Groups
n= 4268
Interested
non-attendees
n= 2061
Percentage
Female 53% 53% 56% 53%
Race/ethnicity
White 78% 81% 81% 70%
Black 14% 12% 12% 20%
Other 8% 7% 7% 10%
Marital status
Married 55% 56% 60% 50%
Separated/divorced/widowed 21% 21% 20% 25%
Never married 24% 23% 20% 25%
Employment status
Employed 63% 65% 62% 57%
Unemployed/not working 6% 5% 6% 7%
Retired 15% 13% 14% 17%
Keeping house 10% 11% 12% 12%
Other 6% 6% 6% 7%
Family income
$0–$9999 18% 17% 16% 28%
$10,000–$19,999 21% 21% 21% 24%
$20,000–$29,999 18% 20% 17% 16%
$30,000–$49,999 23% 21% 23% 21%
$50,000+ 20% 21% 23% 11%
Satisfaction with financial situation
Not satisfied at all 28% 27% –34%
More or less satisfied 44% 44% –44%
Pretty well satisfied 28% 29% –22%
Social class
Lower class 7% 5% 6% 13%
Working class 46% 45% 43% 54%
Middle class 44% 46% 48% 32%
Upper class 3% 4% 3% 1%
General health rating
Excellent 28% 32% –21%
Good 47% 47% –42%
Fair 19% 16% –28%
Poor 6% 5% –9%
Level of urbanicity
Med-large city (50,000+) 31% 31% 29% 31%
Suburb 35% 36% 33% 29%
Unincorporated area 13% 9% 15% 18%
Small city or town 11% 14% 11% 9%
Smaller areas or country 10% 10% 12% 13%
Feels afraid in neighborhood 34% 38% –31%
Bone et al. BMC Public Health (2021) 21:1349 Page 5 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
associated with 1.19 times higher odds (95% CI = 1.16–
1.22) of attendance. Parental education was similarly as-
sociated with increased odds of attendance, although the
estimated odds ratio was smaller (OR = 1.05, 95% CI =
1.04–1.07).
Two residential factors were associated with attend-
ance. Compared to those living in medium to large cit-
ies, the odds of attendance reduced with decreasing level
of urbanicity. The odds of attendance were lowest in
smaller areas or open country. Additionally, for each
additional person in the household, participants had 5%
lower odds (95% CI = 0.90–0.99) of attendance. Partici-
pants who rated their health as fair (OR = 0.68, 95% CI =
0.56–0.83) or poor (OR = 0.47, 95% CI = 0.33–0.66) had
lower odds of attending events than participants who
rated their health as excellent.
Finally, the results suggested that event attendance
varied across survey years, although there was no clear
time trend. In comparison to 1993, the odds of attend-
ance were higher in 1998, 2002, and 2012 but did not
differ in 2016.
Arts activities
Overall, 4372 individuals reported whether they had par-
ticipated in arts activities. These individuals ranged in age
from 18 to 89 years, with a mean age of 44.8 (SD = 17.0).
About 53% were female and 81% were White (Table 1).
On average, 54% reported participating in at least one arts
activity in the last 12 months, and this was relatively stable
across time (1993: 55%, 1998: 51%, 2002: 55%).
Fewer factors were associated with participation in arts
activities than with attendance at arts events (Table 2).
Females had 1.71 times higher odds (95% CI = 1.45–
2.00) of participating than males. Both Black (OR = 0.48,
95% CI = 0.38–0.61) and individuals of Other races/eth-
nicities (OR = 0.70, 95% CI = 0.51–0.96) were less likely
to report participating than those who were White. Indi-
viduals who were unemployed or not working had
higher odds of participating than those working (OR =
1.44, 95% CI = 1.06–1.95). As with attending arts events,
increased years of education (OR = 1.08, 95% CI = 1.05–
1.12) and parental education (OR = 1.05, 95% CI = 1.02–
1.07) were both associated with higher odds of partici-
pating in arts activities. There was no evidence that any
other factors were associated with participation.
Creative groups
Membership of creative groups was reported by 4268
participants, who were similar demographically to partic-
ipants who reported other arts outcomes (Table 1).
Membership in creative groups was lower than attend-
ance at events or participation in activities. Overall, 19%
of participants reported being a member of a creative
group, and this may have decreased over time (1993:
20%, 1994: 16%, 2004: 18%, 2010: 17%).
Despite a lower proportion of participants being mem-
bers of creative groups, membership was associated with
similar factors to arts activities (Table 2). Females had
1.33 times higher odds (95% CI = 1.08–1.63) of member-
ship than males. There was also evidence that the odds
of membership increased with more education (OR =
1.15, 95% CI = 1.10–1.20) and parental education (OR =
1.04, 95% CI = 1.01–1.08). In contrast to arts activities,
those who were never married had 1.58 times higher
odds (95% CI = 1.18–2.11) of membership than married
participants and the odds of membership increased with
age (OR = 1.01, 95% CI = 1.00–1.02). Finally, there was
evidence that membership decreased over time, with the
odds decreasing by 32% (95% CI = 0.54–0.87) from 1993
to 2010.
Interested non-attendees
Overall, 2061 participants reported whether there was
an arts event that they had wanted to go to but did not
attend, 29% of whom were interested non-attendees.
The proportion of interested non-attendees remained
consistent across years (2012: 29%, 2016: 30%).
As with attendance at arts events, there was evidence
that being an interested non-attendee was associated
with race/ethnicity and years of education (Table 2).
Other races/ethnicities had lower odds of being an inter-
ested non-attendee than White individuals (OR = 0.56,
Table 1 Demographic characteristics of the samples, with data combined across all included survey years (Continued)
Events
n= 8684
Activities
n= 4372
Groups
n= 4268
Interested
non-attendees
n= 2061
Mean (SE)
Age 46.61 (0.23) 44.80 (0.33) 45.92 (0.34) 49.14 (0.52)
Years of education 13.44 (0.05) 13.20 (0.07) 13.55 (0.06) 12.55 (0.11)
Parental years of education 12.07 (0.06) 11.85 (0.09) 12.11 (0.09) 11.26 (0.15)
Household size 2.85 (0.02) 2.84 (0.03) 2.88 (0.03) 2.94 (0.06)
Note. Results based on 50 multiply imputed data sets. Events includes participants from survey years 1993, 1998, 2002, 2012, and 2016. Activities includes
participants from 1993, 1998, and 2002. Groups includes participants from 1993, 1994, 2004, and 2010. Interested non-attendees includes participants from 2012
and 2016. SE = standard error
Bone et al. BMC Public Health (2021) 21:1349 Page 6 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Table 2 Logistic regression models testing associations between demographic, socioeconomic, residential, and health exposures
and the odds of arts engagement
Model 1: Events
n = 8684
Model 2: Activities
n = 4372
Model 3: Groups
n = 4268
Model 4: Interested
non-attendes
n = 2061
OR 95% CI p OR 95% CI p OR 95% CI p OR 95% CI p
Age 1.01 0.98–1.03 0.629 1.01 0.98–1.03 0.605 1.01 1.00–1.02 0.007 1.00 0.99–1.01 0.850
Age (quadratic) 1.00 1.00–1.00 0.302 1.00 1.00–1.00 0.129 –– – –– –
Female 1.24 1.10–1.39 < 0.001 1.71 1.45–2.00 < 0.001 1.33 1.08–1.63 0.008 1.19 0.90–1.58 0.215
Race/ethnicity
White 1 1 1 1
Black 0.66 0.55–0.78 < 0.001 0.48 0.38–0.61 < 0.001 0.94 0.66–1.33 0.718 0.92 0.61–1.39 0.696
Other 0.89 0.71–1.11 0.294 0.70 0.51–0.96 0.028 1.11 0.74–1.69 0.606 0.56 0.35–0.89 0.015
Marital status
Married 1 1 1 1
Separated 1.16 1.00–1.34 0.056 0.90 0.75–1.09 0.282 0.91 0.68–1.22 0.531 1.08 0.75–1.57 0.671
Never married 1.37 1.14–1.63 0.001 1.00 0.78–1.27 0.987 1.58 1.18–2.11 0.002 1.26 0.86–1.86 0.236
Employment status
Employed 1 1 1 1
Unemployed 0.93 0.73–1.19 0.561 1.44 1.06–1.95 0.021 0.79 0.50–1.23 0.286 1.41 0.85–2.34 0.179
Retired 1.15 0.93–1.43 0.200 1.10 0.83–1.45 0.523 1.26 0.85–1.85 0.249 0.89 0.57–1.39 0.605
Keeping house 0.82 0.67–1.01 0.057 1.13 0.87–1.46 0.350 1.38 0.95–2.00 0.089 0.95 0.62–1.47 0.831
Other 1.07 0.82–1.39 0.633 1.35 0.96–1.89 0.085 1.11 0.71–1.72 0.651 0.78 0.48–1.28 0.324
Family income
$0–$9999 1 1 1 1
$10,000–$19,999 1.27 1.07–1.51 0.007 0.84 0.65–1.07 0.162 1.15 0.77–1.74 0.492 1.13 0.76–1.68 0.536
$20,000–$29,999 1.58 1.29–1.95 < 0.001 0.96 0.72–1.27 0.756 1.54 0.97–2.45 0.067 1.11 0.66–1.87 0.698
$30,000–$49,999 1.80 1.46–2.22 < 0.001 0.87 0.65–1.17 0.358 1.42 0.91–2.23 0.122 1.14 0.73–1.78 0.557
$50,000+ 2.78 2.17–3.57 < 0.001 0.81 0.58–1.13 0.211 1.42 0.89–2.26 0.137 1.03 0.49–2.15 0.940
Financial situation
Not satisfied at all 1 1 –– – 1
More or less satisfied 0.92 0.80–1.06 0.267 1.00 0.83–1.21 0.962 –– – 0.70 0.52–0.96 0.028
Pretty well satisfied 1.03 0.87–1.21 0.772 1.00 0.81–1.24 0.978 –– – 0.79 0.53–1.17 0.234
Social class
Lower class 1 1 1 1
Working class 1.20 0.94–1.53 0.145 1.20 0.86–1.69 0.285 1.21 0.64–2.30 0.558 1.02 0.65–1.62 0.916
Middle class 1.52 1.16–1.97 0.002 1.03 0.73–1.46 0.870 1.35 0.71–2.57 0.359 0.69 0.43–1.12 0.132
Upper class 1.52 0.99–2.35 0.058 0.92 0.56–1.50 0.743 1.67 0.77–3.62 0.195 0.29 0.08–1.09 0.066
Years of education 1.19 1.16–1.22 < 0.001 1.08 1.05–1.12 < 0.001 1.15 1.10–1.20 < 0.001 1.11 1.05–1.17 < 0.001
Parental years of education 1.05 1.04–1.07 < 0.001 1.05 1.02–1.07 < 0.001 1.04 1.01–1.08 0.019 1.03 0.99–1.08 0.147
General health rating
Excellent 1 1 –– – 1
Good 0.88 0.75–1.03 0.121 0.95 0.79–1.14 0.577 –– – 1.02 0.67–1.55 0.917
Fair 0.70 0.58–0.85 < 0.001 0.92 0.71–1.18 0.492 –– – 1.30 0.84–2.02 0.243
Poor 0.48 0.34–0.67 < 0.001 0.91 0.63–1.33 0.634 –– – 1.38 0.74–2.56 0.309
Level of urbanicity
Med-large city 1 1 1 1
Bone et al. BMC Public Health (2021) 21:1349 Page 7 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
95% CI = 0.35–0.89), and the odds of interested non-
attendance increased with level of education (OR =
1.11, 95% CI = 1.05–1.17). However, in contrast to
event attendance, those who were more or less satis-
fied with their financial situation had lower odds of
being an interested non-attendee than those who were
not satisfied at all (OR = 0.70, 95% CI = 0.52–0.96).
There was no evidence that being an interested non-
attendee was associated with gender, marital status,
employment status, family income, social class, paren-
tal education, level of urbanicity, household size, or
general health rating.
Table 2 Logistic regression models testing associations between demographic, socioeconomic, residential, and health exposures
and the odds of arts engagement (Continued)
Model 1: Events
n = 8684
Model 2: Activities
n = 4372
Model 3: Groups
n = 4268
Model 4: Interested
non-attendes
n = 2061
OR 95% CI p OR 95% CI p OR 95% CI p OR 95% CI p
Suburb 0.92 0.79–1.08 0.311 1.18 0.98–1.42 0.075 1.06 0.82–1.38 0.666 1.12 0.77–1.63 0.558
Unincorporated area 0.79 0.64–0.96 0.020 0.96 0.75–1.24 0.771 1.22 0.89–1.67 0.225 1.15 0.77–1.73 0.495
Small city or town 0.69 0.58–0.83 < 0.001 1.08 0.84–1.41 0.538 1.20 0.88–1.64 0.255 1.19 0.77–1.85 0.436
Smaller areas 0.57 0.47–0.69 < 0.001 0.98 0.75–1.28 0.874 0.96 0.63–1.45 0.836 0.63 0.39–1.02 0.061
Household size 0.95 0.90–0.99 0.030 1.02 0.95–1.09 0.572 0.99 0.91–1.08 0.811 0.96 0.87–1.05 0.350
Feels afraid in neighborhood 1.06 0.91–1.24 0.463 0.97 0.80–1.17 0.714 –– – 1.11 0.79–1.56 0.536
Survey year
11111
22.02 1.69–2.40 < 0.001 0.89 0.74–1.06 0.194 0.73 0.52–1.03 0.074 1.04 0.79–1.38 0.757
32.27 1.88–2.73 < 0.001 1.05 0.88–1.26 0.568 0.77 0.60–0.99 0.045 –– –
41.25 1.06–1.48 0.008 –– – 0.68 0.54–0.87 0.002 –– –
5 1.05 0.87–1.26 0.624 –– – –– – –– –
Note. Survey year refers to different years for each arts outcome: for events 1 =1993, 2 = 1998, 3 = 2002, 4 = 2012, 5 = 2016; for activities 1 =1993, 2 =1998, 3 =
2002; for groups 1 = 1993, 2 = 1994, 3 = 2004, 4 = 2010; and for interested non-attendees 1 = 2012, 2 = 2016. These numbers have been added for ease of
presentation; years were used in analyses. For odds ratios, 1 indicates the reference category
Fig. 1 Results of subgroup analyses, with logistic regression models testing associations between exposures and the odds of attending arts
events separately in each survey year (1993 n= 1590, 1998 n= 1432, 2002 n= 1355, 2012 n= 2838, 2016 n= 1469). Odds ratios and 95%
confidence intervals are displayed. For associations between sex and arts events, the odds ratio represents attendance in females compared to
males. For associations between race/ethnicity and arts events, White is the reference category. Associations were estimated in the full logistic
regression models (including all exposures as shown in Table 2), but only results for sex and race/ethnicity are presented
Bone et al. BMC Public Health (2021) 21:1349 Page 8 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Change across survey years
Next, we tested whether associations between arts en-
gagement outcomes and age, sex, race/ethnicity, class,
and income differed over time. There was no evidence
for interactions between survey year and any exposures
on participation in arts activities, membership of creative
groups, or being an interested non-attendee (Supple-
mentary Table 5). There was also no evidence for inter-
actions between survey year and age, class, or income on
arts event attendance.
However, there was evidence for an interaction be-
tween survey year and sex on event attendance. There
was no linear time trend, as females had higher odds of
attendance than males in 1993 and 2002 but there were
no sex differences in other survey years (Fig. 1; Supple-
mentary Table 6). There was also evidence for an inter-
action between survey year and race/ethnicity on event
attendance. Black participants had lower odds of attend-
ing than White participants, and this difference in-
creased over time (Fig. 1; Supplementary Table 6).
Sensitivity analyses
We have reported findings based on imputed data but
the results of analyses did not vary when limited to
complete cases, as shown in Supplementary Table 4.In
our sensitivity analysis, limiting the sample to the most
homogenous definitions of arts event attendance (i.e. ex-
cluding participants from 1993) did not substantially
alter our findings (Supplementary Table 7).
Discussion
In this study, we examined whether there are social in-
equalities in engagement in the arts, as found in other
health-related behaviors [3–5]. Between 1993 and 2016,
approximately half of our sample reported attending arts
events, and a similar proportion participated in arts ac-
tivities. In the smaller sample of individuals who com-
pleted the GSS in 2012 and 2016, another one third
were interested non-attendees, who had been interested
in attending an event in the last year but had not gone
to it. Fewer people were members of creative groups,
with approximately one fifth of the sample between
1993 and 2010 reporting group membership. Several
demographic factors were consistently associated with
engagement in the arts. For example, engagement was
higher in females than males, and married individuals
were less likely to engage than those who had never
married. Attendance at arts events and participation in
arts activities also differed according to race/ethnicity,
although creative group membership did not. Socioeco-
nomic factors showed mixed associations with the differ-
ent types of arts engagement. Higher levels of education
and parental education were consistently associated with
all types of engagement. Attendance at arts events was
also associated with higher income and social class, bet-
ter health, and living in more urban areas. However, be-
ing an interested non-attendee of arts events was not
associated with these factors. In contrast to arts events,
we found no evidence that income, social class, health,
or urbanicity were associated with participation in arts
activities and groups. Most of our findings are consistent
with previous research demonstrating that a number of
demographic and socioeconomic factors are associated
with engagement in the arts [13]. Our findings further
advance previous research by using a broader definition
of arts to more accurately reflect the breadth of engage-
ment in the US.
The associations between several demographic factors,
such as sex and marital status, and engagement in all
forms of the arts are consistent with previous evidence
[19,40–44]. We also found that race/ethnicity was more
strongly associated with participation in arts activities
than events, as shown previously [22]. This association
was independent of socioeconomic factors, so is unlikely
to be explained by over-representation of ethnic minor-
ities in lower socioeconomic status groups [45]. A report
that also used GSS data found that lower attendance at
arts events by racial/ethnic minorities may be a result of
barriers such as being unable to get to the venue and
not having anyone to go with [23]. These individuals
were also more likely to state celebrating their cultural
heritage as a reason for attending events than those who
were White [23,46]. However, in this study, we found
that Other races/ethnicities were also less likely to be in-
terested non-attendees of arts events than White indi-
viduals. Although this could be a result of the way in
which arts events were defined (limited to music, the-
atre, or dance performances or art exhibits), it may also
indicate that some ethnic/racial groups are less inter-
ested in attending arts events. A lack of cultural equity,
cultural relevance, interest, and inequalities in access are
therefore likely to contribute to the racial/ethnic differ-
ences in arts engagement.
Overall, our findings support previous evidence that
education is most strongly associated with engagement
in the arts [12,13,18–24]. However, contrary to some
recent evidence, we did not find that education was
more strongly associated with attending events than
other forms of arts engagement [25]. Education may in-
crease engagement by helping to cultivate cultural tastes
and preferences, raising awareness of activities, and in-
creasing cognitive capacity to engage [47]. Arts educa-
tion specifically may also contribute to this association,
as it is strongly related to both level of education and
arts engagement [20,27,32,48,49]. We found a similar
association with parental education, independent of the
individual’s own education, although the magnitude of
association was smaller. This indicates that childhood
Bone et al. BMC Public Health (2021) 21:1349 Page 9 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
socioeconomic status continues to influence engagement
in the arts throughout the lifecourse. Children of parents
with more education may benefit from increased access
to the arts during development and may be more likely
to receive arts education in childhood (e.g. learning to
play an instrument [30]). These individuals may there-
fore have more training and experience, enabling them
to participate in more highly skilled arts activities (e.g.
orchestras).
Consistent with previous evidence for a social gradient
in arts engagement, we found that attendance at arts
events was less likely with lower income and social class,
poorer health, and less urban areas. As being an inter-
ested non-attendee was not associated with these factors,
they are likely to be barriers specifically to attendance.
Individuals across the range of incomes, social classes,
health, and levels of urbanicity were interested in attend-
ing events at similar rates, but actual attendance differed
according to these factors. Previously, individuals with
lower household income and social class were more
likely to report barriers to attending events of cost and
difficulty of getting to a venue, as well as a lack of time
[23,46]. Other research has demonstrated that individ-
uals with poorer physical health may experience more
barriers affecting their perceived capabilities to engage
[30]. Areas that are more urban, such as cities, are likely
to have a larger range of arts events on offer, including
at a variety of times and costs as well as appealing to a
broader audience, and events may be more geographic-
ally dispersed or easier to attend using public transport.
Urbanicity can thus be interpreted as a proxy measure
for the availability of arts events. However, there are also
likely to be area-level factors related to the availability
and accessibility of the arts that, although not measured
in the GSS, require further investigation. In contrast to
arts events, we found no evidence that income, social
class, health, or urbanicity were associated with partici-
pation in arts activities and groups. These types of en-
gagement may be more widely available, include more
diverse activities, be cheaper to participate in, and often
do not require attendance at a specific venue, which may
be hard to reach or not generally attended by certain
groups.
There was some mixed evidence for a social gradient
in interest in arts events. Individuals with higher levels
of education were more likely to be interested non-
attendees, as were people who were more or less satis-
fied with their financial situation (compared to those
who were not satisfied at all). Previous research has sug-
gested that of the different types of arts engagement,
education is most strongly associated with highbrow cul-
tural events [25], which could explain the association
with interest in events. It is unclear why we found evi-
dence for an association with financial satisfaction. We
might conclude that individuals who were satisfied with
their financial situation were not interested non-
attendees because they were financially able to attend
any events of interest, but we found no evidence that fi-
nancial satisfaction was associated with actual event at-
tendance. Additionally, there was no evidence that being
an interested non-attendee was associated with income
or differed between those who were pretty well satisfied
and not at all satisfied with their financial situation. The
relationship between interest in the arts, subjective mea-
sures of satisfaction with financial situation, and more
objective measures of income thus requires further
investigation.
We also investigated changing patterns of arts engage-
ment as there has been concern that arts participation is
decreasing in the US [11,22,32]. We found some evi-
dence that event attendance changed over time, but this
was likely a result of changes in the measure of event at-
tendance, as there was no linear trend. In contrast,
group membership decreased over time. Additionally,
the racial disparity in event attendance, with an over-
representation of White individuals compared to those
of racial/ethnic minorities, increased from 1993 to 2016.
These increasing racial/ethnic inequalities in arts event
attendance were independent of other socioeconomic
factors such as income and education. However, given
the nature of structural racism, this finding should be
interpreted cautiously and requires replication in studies
with consistent measures of event attendance. As this
study spanned a period of 23 years, with event attend-
ance and group membership measured at different
times, specific social and economic events in each year
could also have contributed to the changing patterns of
arts engagement.
Our findings have implications for understanding
health and social inequalities in the US. A number of the
factors that we have identified as associated with arts en-
gagement are also associated with inequalities in access
to health care and health outcomes [14–17]. This could
be because arts engagement is a correlate of health, with
both representing a form of capital that can be obtained
by individuals with more material resources, such as in-
come, and non-material resources, such as social sup-
port [47]. Consistent with this, we found evidence that
poorer self-reported health was associated with lower at-
tendance at arts events, although it was not associated
with interest in attending events or participation in arts
activities. Arts engagement could also represent a health
behaviour that leads to improved health outcomes.
There is growing evidence that engagement with the arts
can lead to a range of health benefits, independent of
demographic and socioeconomic factors [9,50]. It is
thus concerning that we have found evidence for differ-
ential engagement in the arts. Future research should
Bone et al. BMC Public Health (2021) 21:1349 Page 10 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
explore why engagement is lower in these groups, in
particular males, racial/ethnic minorities, and those with
lower education. This is particularly important given
that previous efforts to reduce inequalities in access to
cultural events by expanding facilities and offering free
tickets in Brazil have not been successful [51]. Future re-
search could also investigate whether removing other
barriers to engagement, such as providing the arts online
to avoid high prices and reduce time constraints, could
increase levels of engagement [52]. This could then sup-
port the development of interventions to promote en-
gagement in the arts, and test whether this leads to
improvements in health outcomes.
This study has a number of strengths. The GSS was a
large nationally representative sample and we included
several measures of arts engagement. Although the GSS
has previously been used to study arts engagement [23,
43], research has not generally examined membership of
creative groups in comparison to other forms of engage-
ment or combined data across as many waves of the GSS
as in this study. We tested a range of factors that may be
associated with arts engagement, and mutually adjusted
for these variables in our models. Using multiple imput-
ation means that missing data should not have influenced
our findings. However, this study also has a number of
limitations. We tested cross-sectional associations and
thus cannot rule out the possibility of inverse causality.
There are some factors, such as health, which may have a
bidirectional association with arts engagement. Addition-
ally, the GSS did not measure attendance at arts events
consistently across waves, which is likely to explain the as-
sociation we found between event attendance and survey
year. A broader definition of arts events was used in later
years. However, when limiting our analyses just to this
broader definition, our findings were consistent. Although
our measures of arts engagement were more inclusive
than in many previous studies, they were likely still too
narrow. Standard arts engagement questions are not able
to capture arts engagement in some immigrant communi-
ties [35], and also typically do not cover engagement in
digital or electronic arts activities such as graphic design,
photography, film-making, and music production. This
could have contributed to our findings of lower arts en-
gagement in individuals who were not White and under-
represented arts engagement amongst younger genera-
tions. Future research should aim to measure diverse as-
pects of arts engagement, particularly as the US moves
towards a majority-minority society, in which the non-
Hispanic white population will no longer form the major-
ity of the US population [53].
Conclusions
Given the potential importance of engagement in the
arts for health and wellbeing [9], individuals should be
provided with equal opportunities to participate. Our
findings indicate that social determinants may influence
engagement in the arts throughout the life course. En-
couraging arts activities and creative group membership
may provide one way of widening participation and re-
ducing social inequalities in arts engagement. It will also
be important to recognize that lack of participation may
not merely be due to a lack of interest or motivation but
may be influenced by structural barriers, such as racism,
or a lack of opportunities. Indeed, the nature of many
arts activities that take place in well defined arts spaces
are rooted in white supremacy, creating a foundational
barrier for Black, Indigeouns and other people of color
(BIPOC) groups. Future research is needed to identify
what these barriers are and how they can be removed.
This is particularly important in the wake of COVID-19,
given the closure of many arts venues and the dispropor-
tionate effect on BIPOC individuals and those of lower
socioeconomic status [28,54–56].
Supplementary Information
The online version contains supplementary material available at https://doi.
org/10.1186/s12889-021-11263-0.
Additional file 1.
Acknowledgements
We thank Shanae Burch, Nupur Chaudhury, and David Fakunle, thought
leaders on work at the intersections of the arts, equity, and public health in
the US, for their comments on this manuscript. We also gratefully
acknowledge the contribution of the GSS study participants.
Authors’contributions
JKB, FB, and DF designed the study. JKB conducted the analysis and drafted
the manuscript. JKB, FB, MEF, EP, JKS, and DF contributed to the writing,
made critical revisions, and approved the final manuscript.
Funding
The EpiArts Lab, a National Endowment for the Arts Research Lab at the
University of Florida, is supported in part by an award from the National
Endowment for the Arts (1862896–38-C-20). The opinions expressed are
those of the authors and do not represent the views of the National
Endowment for the Arts Office of Research & Analysis or the National
Endowment for the Arts. The National Endowment for the Arts does not
guarantee the accuracy or completeness of the information included in this
material and is not responsible for any consequences of its use. The EpiArts
Lab is also supported by the University of Florida, the Pabst Steinmetz
Foundation, and Bloomberg Philanthropies. DF is supported by the
Wellcome Trust [205407/Z/16/Z].
Availability of data and materials
The dataset supporting the conclusions of this article is available in the GSS
repository, https://gss.norc.org/get-the-data/stata.
Declarations
Ethics approval and consent to participate
All GSS participants gave informed consent and this study has Institutional
Review Board approval from the University of Florida (IRB201901792) and
ethical approval from University College London Research Ethics Committee
(project 18839/001). All methods were carried out in accordance with
relevant ethical guidelines and regulations, the Helsinki Declaration (2013
revision), and the General Data Protection Regulation.
Bone et al. BMC Public Health (2021) 21:1349 Page 11 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Consent for publication
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Author details
1
Research Department of Behavioural Science and Health, Institute of
Epidemiology & Health, University College London, London, UK.
2
Center for
Arts in Medicine, University of Florida, Gainesville, Florida, USA.
Received: 24 February 2021 Accepted: 9 June 2021
References
1. Bleich SN, Jarlenski MP, Bell CN, Laveist TA. Health inequalities: trends,
progress, and policy. Annu Rev Public Health. 2012;33(1):7–40. https://doi.
org/10.1146/annurev-publhealth-031811-124658.
2. Zaninotto P, Batty GD, Stenholm S, Kawachi I, Hyde M, Goldberg M, et al.
Socioeconomic inequalities in disability-free life expectancy in older people
from England and the United States: a cross-national population-based
study. J Gerontol Ser A Biol Sci Med Sci. 2020;75(5):906–13. https://doi.org/1
0.1093/gerona/glz266.
3. Stringhini S, Sabia S, Shipley M, Brunner E, Nabi H, Kivimaki M, et al.
Association of socioeconomic position with health behaviors and mortality.
J Am Med Assoc. 2010;303(12):1159–66. https://doi.org/10.1001/jama.2010.2
97.
4. Harper S, Lynch J. Trends in socioeconomic inequalities in adult health
behaviors among U.S. states, 1990-2004. Public Health Rep. 2007;122(2):177–
89. https://doi.org/10.1177/003335490712200207.
5. Scholes S, Bann D. Education-related disparities in reported physical activity
during leisure-time, active transportation, and work among US adults:
repeated cross-sectional analysis from the National Health and nutrition
examination surveys, 2007 to 2016. BMC Public Health. 2018;18:1–10.
https://doi.org/10.1186/s12889-018-5857-z.
6. Singh-Manoux A, Marmot M. Role of socialization in explaining social
inequalities in health. Soc Sci Med. 2005;60(9):2129–33. https://doi.org/10.1
016/j.socscimed.2004.08.070.
7. Mak HW, Coulter R, Fancourt D. Patterns of social inequality in arts and
cultural participation: findings from a nationally representative sample of
adults living in the United Kingdom of Great Britain and Northern Ireland.
Public Heal Panor. 2020;6:55–68.
8. Lamont M, Beljean S, Clair M. What is missing? Cultural processes and causal
pathways to inequality. Socio Econ Rev. 2014;12(3):573–608. https://doi.
org/10.1093/ser/mwu011.
9. Fancourt D, Finn S. What is the evidence on the role of the arts in
improving health and well-being? A scoping review. Copenhagen; 2019.
https://europepmc.org/article/NBK/nbk553773
10. Fancourt D, Aughterson H, Finn S, Walker E, Steptoe A. How leisure activities
affect health: A review and multi-level theoretical framework of mechanisms
of action using the lens of complex adaptive systems science. Lancet
Psychiatry. 2021. https://doi.org/10.1016/S2215-0366(20)30384-9.
11. National Endowment for the Arts. How a Nation Engages With Art:
Highlights from the 2012 Survey of Public Participation in the Arts.
Washington, DC; 2013. https://www.arts.gov/sites/default/files/highlights-
from-2012-sppa-revised-oct-2015.pdf
12. National Endowment for the Arts. U.S. Patterns of Arts Participation: A Full
Report from the 2017 Survey of Public Participation in the Arts. Washington,
DC; 2019. https://www.arts.gov/sites/default/files/US_Patterns_of_Arts_Pa
rticipationRevised.pdf
13. Stallings SN, Mauldin B. Public engagement in the arts: a review of recent
literature. 2016. https://www.lacountyarts.org/sites/default/files/pdfs/lacac_
pubenglitrev.pdf.
14. Mays VM, Cochran SD, Barnes NW. Race, race-based discrimination, and
health outcomes among African Americans. Annu Rev Psychol. 2007;58(1):
201–25. https://doi.org/10.1146/annurev.psych.57.102904.190212.
15. Nguyen AB, Moser R, Chou WY. Race and health profiles in the United
States: an examination of the social gradient through the 2009 CHIS adult
survey. Public Health. 2014;128(12):1076–86. https://doi.org/10.1016/j.puhe.2
014.10.003.
16. Williams DR, Priest N, Anderson NB. Understanding associations among
race, socioeconomic status, and health: patterns and prospects. Health
Psychol. 2016;35(4):407–11. https://doi.org/10.1037/hea0000242.
17. Singh GK. Area deprivation and widening inequalities in US mortality, 1969-
1998. Am J Public Health. 2003;93(7):1137–43. https://doi.org/10.2105/AJPH.
93.7.1137.
18. Robinson JP. Arts participation in America: 1982–1992. 1993. https://www.a
rts.gov/sites/default/files/NEA-Research-Report-27.pdf.
19. Peterson RA, Hull PC, Kern RM. Age and arts participation: 1982–1997. 2000.
https://www.arts.gov/sites/default/files/ResearchReport42.pdf.
20. Ostrower F. The diversity of cultural participation: Findings from a national
survey; 2005. https://doi.org/10.4324/9780203927502.
21. Seaman BA. Empirical Studies of Demand for the Performing Arts. In V. A.
Ginsburg & D. Throsby (Eds.), Handbook of the Economics of Art and
Culture. 2006;1:415–72. https://doi.org/10.1016/S1574-0676(06)01014-3.
22. Welch V, Kim Y. Race/ethnicity and arts participation: findings from the
survey of public participation in the arts. 2010. https://files.eric.ed.gov/
fulltext/ED519762.pdf.
23. Blume-Kohout M, Leonard SR, Novak-Leonard JL. When Going Gets Tough:
Barriers and Motivations Affecting Arts Attendance. 2015. https://www.arts.
gov/sites/default/files/when-going-gets-tough-revised2.pdf.
24. Ateca-Amestoy V, Gorostiaga A, Rossi M. Motivations and barriers to
heritage engagement in Latin America: tangible and intangible dimensions.
J Cult Econ. 2020;44(3):397–423. https://doi.org/10.1007/s10824-019-09366-z.
25. Suarez-Fernandez S, Prieto-Rodriguez J, Perez-Villadoniga MJ. The changing
role of education as we move from popular to highbrow culture. J Cult
Econ. 2020;44(2):189–212. https://doi.org/10.1007/s10824-019-09355-2.
26. DiMaggio P, Ostrower F. Race, ethnicity and participation in the arts.
Washington, D.C.; 1992. https://www.arts.gov/sites/default/files/NEA-Resea
rch-Report-25.pdf
27. Borgonovi F. Performing arts attendance: an economic approach. Appl
Econ. 2004;36(17):1871–85. https://doi.org/10.1080/0003684042000264010.
28. Egede LE, Walker RJ. Structural racism, social risk factors, and Covid-19 - a
dangerous convergence for black Americans. N Engl J Med. 2020;383(12):
e77. https://doi.org/10.1056/NEJMp2023616.
29. Mak HW, Coulter R, Fancourt D. Does arts and cultural engagement vary
geographically? Evidence from the UK household longitudinal study. Public
Health. 2020;185:119–26. https://doi.org/10.1016/j.puhe.2020.04.029.
30. Fancourt D, Mak HW. What barriers do people experience to engaging in
the arts? Structural equation modelling of the relationship between
individual characteristics and capabilities, opportunities, and motivations to
engage. PLoS One. 2020;15(3):1–12. https://doi.org/10.1371/journal.pone.023
0487.
31. Novak-Leonard JL, Brown AS, Brown W. Beyond attendance: a multi-modal
understanding of arts participation. 2011. https://www.arts.gov/sites/default/
files/2008-SPPA-BeyondAttendance.pdf.
32. Rabkin N, Hedberg EC. Arts education in America: what the declines mean
for arts participation. 2012. https://www.arts.gov/sites/default/files/2008-
SPPA-ArtsLearning.pdf.
33. Jackson M-R, Herranz J, Kabwasa-Green F. Art and culture in communities:
unpacking participation. Washington, DC; 2003. http://webarchive.urban.
org/UploadedPDF/311006_unpacking_participation.pdf
34. Cowen T. In praise of commercial culture: Harvard University Press; 1998.
35. Novak-Leonard JL, O’Malley MK, Truong E. Minding the gap: elucidating the
disconnect between arts participation metrics and arts engagement within
immigrant communities. Cult Trends. 2015;24(2):112–21. https://doi.org/10.1
080/09548963.2015.1031477.
36. Smith TW, Davern M, Freese J, Morgan S. General Social Surveys, 1972–2018
[machine-readable data file]: Sponsored by National Science Foundation.
NORC at the University of Chicago [producer and distributor]. Data accessed
from the GSS Data Explorer website at gssdataexplorer.norc.org; 2019.
37. Thompson CG, Kim RS, Aloe AM, Becker BJ. Extracting the variance in flation
factor and other multicollinearity diagnostics from typical regression results.
Basic Appl Soc Psychol. 2017;39(2):81–90. https://doi.org/10.1080/01973533.2
016.1277529.
38. White IR, Royston P, Wood AM. Multiple imputation using chained
equations: issues and guidance for practice. Stat Med. 2011;30(4):377–99.
https://doi.org/10.1002/sim.4067.
39. StataCorp. Stata Statistical Software: Release 16. 2019.
40. DiMaggio P. Gender, networks, and cultural capital. Poetics. 2004;32(2):99–
103. https://doi.org/10.1016/j.poetic.2004.02.004.
Bone et al. BMC Public Health (2021) 21:1349 Page 12 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
41. Christin A. Gender and highbrow cultural participation in the United States.
Poetics. 2012;40(5):423–43. https://doi.org/10.1016/j.poetic.2012.07.003.
42. Schmutz V, Stearns E, Glennie EJ. Cultural capital formation in adolescence:
high schools and the gender gap in arts activity participation. Poetics. 2016;
57:27–39. https://doi.org/10.1016/j.poetic.2016.04.003.
43. Lewis GB, Seaman BA. Sexual orientation and demand for the arts. Soc Sci
Q. 2004;85(3):523–38. https://doi.org/10.1111/j.0038-4941.2004.00231.x.
44. Montgomery SS, Robinson MD. Empirical evidence of the effects of
marriage on male and female attendance at sports and arts. Soc Sci Q.
2010;91(1):99–116. https://doi.org/10.1111/j.1540-6237.2010.00683.x.
45. Semega J, Kollar M, Shrider EA, Creamer JF. Income and Poverty in the
United States: 2019. US Census Bur Curr Popul Reports. 2020;September:60–
270.
46. Dwyer C, Weston-Sementelli J, Lewis CK, Hiebert-Larson J. Why we engage:
attending, creating, and performing art. 2020. https://www.arts.gov/sites/
default/files/Why-We-Engage-0920_0.pdf.
47. Bourdieu P. The forms of capital. In: Richardson J, editor. Handbook of
theory and research for the sociology of education. New York: Greenwood
Press; 1986. p. 241–58.
48. Novak-Leonard JL, Baach P, Schultz A, Farrell B, Anderson W, Rabkin N. The
Changing Landscape of Arts Participation: A Synthesis of Literature and
Expert Interviews. 2014. https://doi.org/10.6082/uchicago.1262.
49. Elpus K. Music education promotes lifelong engagement with the arts.
Psychol Music. 2018;46(2):155–73. https://doi.org/10.1177/030573561
7697508.
50. Fancourt D, Steptoe A. Cultural engagement and mental health: Does
socio-economic status explain the association? Soc Sci Med. 2019;
236(February):112425. https://doi.org/10.1016/j.socscimed.2019.112425.
51. de Almeida CCR, Lima JPR, Gatto MFF. Expenditure on cultural events:
preferences or opportunities? An analysis of Brazilian consumer data. J Cult
Econ. 2020;44(3):451–80. https://doi.org/10.1007/s10824-019-09370-3.
52. De la Vega P, Suarez-Fernández S, Boto-García D, Prieto-Rodríguez J. Playing
a play: online and live performing arts consumers profiles and the role of
supply constraints. J Cult Econ. 2020;44(3):425–50. https://doi.org/10.1007/s1
0824-019-09367-y.
53. U.S. Census Bureau. 2017 National Population Projections Tables 2018.
https://www.census.gov/data/tables/2017/demo/popproj/2017-summary-ta
bles.html.
54. Brown CS, Ravallion M. Inequality and the coronavirus: socioeconomic
covariates of behavioral responses and viral outcomes across US counties.
2020. http://www.nber.org/papers/w27549.
55. van Dorn A, Cooney RE, Sabin ML. COVID-19 exacerbating inequalities in
the US. Lancet. 2020;395(10232):1243–4. https://doi.org/10.1016/S0140-673
6(20)30893-X.
56. Bowleg L. We’re not all in this together: on COVID-19, intersectionality, and
structural inequality. Am J Public Health. 2020;110(7):917–8. https://doi.org/1
0.2105/AJPH.2020.305766.
Publisher’sNote
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Bone et al. BMC Public Health (2021) 21:1349 Page 13 of 13
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Available via license: CC BY 4.0
Content may be subject to copyright.