Investigating Socioeconomic Disparities in Park Access Across Two Southeastern US Cities Using a Unique Park Metric: An Ecological Analysis

Abstract

Purpose

This study aimed to demonstrate how ParkIndex values can be used to examine issues related to composite park access, and to investigate whether ParkIndex scores for block groups were significantly different by income, race/ethnicity, and population density.

Design

Cross-sectional, environmental justice analysis of spatial disparities.

Setting

Greenville County, South Carolina, and Raleigh, North Carolina.

Subjects

Park access and quality data were collected for a total of 507 block groups across the study setting.

Measures

For all study block groups, we empirically derived a multi-dimensional park access metric called ParkIndex representing the probability of park use based on three park availability and quality indicators. Block group median household income, race/ethnicity, and population density were obtained from the US Census Bureau’s American Community Survey.

Analysis

Choropleth maps were used to juxtapose ParkIndex values and neighborhood sociodemographic attributes. Analysis of variance was conducted to examine differences in ParkIndex values by multiple sociodemographic variables.

Results

No significant differences in ParkIndex values were found by median household income (F = 0.15,P = .86), percent non-Hispanic White race/ethnicity (F = 0.22,P = .80), or population density (F = 0.70,P = .50).

Conclusion

The current study contributes to the fast-growing body of literature examining whether access to quality parks is equitably distributed in diverse locations. Future research can investigate the utility and generalizability of this composite metric to identify and mitigate potential environmental injustices and health disparities.

Keywords

parks socioeconomic status race/ethnicity disparities measurement

Purpose

Parks are recognized as key components of healthy communities for the diverse health, social, economic, and environmental benefits they provide.^1-3 They offer valuable social and physical activity settings and have been linked with important outcomes such as reduced childhood obesity and lesser rates of chronic disease.^4-6 However, several studies have documented that access to parks is often not equitably distributed between and within neighborhoods and communities.^7-12 For example, important elements of parks, such as resource availability, amenities, facilities, maintenance, and quality, can differ according to various neighborhood socioeconomic factors such as income and race/ethnicity.^7,9,12 However, such studies have often been limited in that they only examine one element of park access, such as the amount of greenspace, specific facilities like playgrounds, or elements of quality.^1,13-17 In contrast, all of these components are important for influencing park use and residents’ health, yet minimal research has been conducted to examine a composite metric of access to quality parks and how this may differ according to important neighborhood attributes.

The ParkIndex study developed an empirical multi-dimensional park access metric for use by diverse communities of users.^18-20 The study leveraged an innovative map-based survey system (Maptionnaire; https://maptionnaire.com/) to collect residents’ reports of visiting specific parks within their neighborhood located across 128 total block groups in four diverse cities.¹⁹ The final ParkIndex formula was composed of three key variables associated with respondents’ use of neighborhood parks: number of parks within one half-mile, total park acreage within one half-mile, and average park quality score.¹⁹ ParkIndex was calculated using values for the significant park access variables in the model multiplied by their respective coefficients and represents the probability of park use (0-100) for a given point/address. To date, ParkIndex has been used to identify areas in need of intervention^18,19 and to understand predictors of park-based physical activity.²¹ However, an essential contribution of ParkIndex may also be to elucidate disparities in composite park availability and quality across neighborhoods with varying sociodemographic attributes (i.e., park deserts). This is important for environmental justice in documenting unequal access to health-promoting resources and in providing an easy to understand metric and tool for use by researchers, citizens, planners, and other parties concerned with park equity.^22-24

Given these considerations, this study had two main objectives: (1) to demonstrate how ParkIndex values can be used to examine issues related to composite park access, and (2) to examine whether ParkIndex scores for block groups were significantly different by income, race/ethnicity, and population density.

Methods

Design and Setting

This study employed a cross-sectional design. As is described further elsewhere,^18,19 the broader ParkIndex study was conducted in select areas (32 census block groups each) of four diverse cities in the United States – Brooklyn, NY, Greenville County, SC, Raleigh, NC, and Seattle, WA. However, the present analysis focuses only on Greenville County, SC, and Raleigh, NC, where data collection about park access was expanded to all block groups in both areas (N = 255 in Greenville County, N = 252 in Raleigh) to ensure a complete and representative sample for studying issues related to park equity. Both study areas are southern, moderately sized metropolitan areas containing a mixture of urban and suburban space and an almost identical number of block groups. At the time of the study, the populations of Greenville County and Raleigh were 499 122 and 441 326 inhabitants, respectively, while 120 and 133 parks were included in the study for each area, respectively.

Measures

For all block groups, data were downloaded from the US Census Bureau’s American Community Survey (5-year estimates, 2011-2015).²⁵ Median household income was categorized into three similar-sized tertiles (low, intermediate, high) based on the distribution of block groups within each city. Race/ethnicity was calculated as the percentage of non-Hispanic White population (all other groups were aggregated due to limited proportions of other races/ethnicities) and was also similarly categorized into tertiles. Finally, population density (total population/area) was calculated for each block group and categorized into low, intermediate, and high within each city. Block group age structure was also captured as the percentage of population less than 18 years.

ParkIndex values were calculated for all block groups in both cities. ParkIndex is an innovative metric composed of three key park variables – number of parks, total park acreage, and an average park quality score for all parks within a one half-mile network buffer. The latter component was captured using the electronic Community Park Audit Tool^26,27 and was comprised of six key components: park access amenities (e.g., adjacent sidewalk), park facilities (e.g., tennis court), park quality and safety amenities (e.g., restroom), park aesthetic features (e.g., historical feature), park quality concerns (e.g. litter), and neighborhood quality concerns (e.g., heavy traffic); the latter two variables reverse-scored).¹⁹ ParkIndex values can be calculated for any address or point and were derived for the centroid of all 100 m × 100 m cells across both Greenville County and Raleigh.¹⁹ All such cell values within each block group were then averaged to calculate a ParkIndex value for all block groups.

Analyses

ParkIndex values were mapped for all block groups in both cities and overlayed with block group values for income, percent non-Hispanic White race/ethnicity, and population density. One-way analysis of covariance (ANCOVA) was used to examine differences in ParkIndex values according to tertiles (low, intermediate, high) for all three block group attributes. For all analyses, block group-level age structure (i.e., percentage of population <18 years, categorized into tertiles) and location (Greenville County or Raleigh) were included as covariates. All analyses were conducted in SAS 9.4 (Cary, NC), and tests were considered significant at P < .05.

Results

A total of 507 block groups were included in our sample, distributed across Greenville (n = 255, 50.3%) and Raleigh (n = 252, 49.7%). Table 1 presents values for block groups classified as low, intermediate, or high median household income, percentage non-Hispanic White population, and population density in both locations. In addition, mean ParkIndex values were 44.6 (SD = 33.1) in Greenville County and 73.0 (SD = 29.0) in Raleigh (Table 1).

Table 1.

Characteristics of Study Area Block Groups (N = 507)

Block groupCharacteristics	Greenville		Raleigh
Block groupCharacteristics	N	Range or mean (SD)	N	Range or mean (SD)
	255		252
Median household income tertiles ($)
Low	85	0-38,690	84	0-42,206
Intermediate	85	38 691-57,250	84	42 207-69,712
High	85	57 251-151,630	84	69 713-183,836
Percent White race tertiles (%)
Low	85	3.5-60.5	84	0.5-47.0
Intermediate	85	60.6-82.3	84	47.1-76.1
High	85	82.4-99.4	84	76.2-100.0
Population density tertiles (population/km²)
Low	85	5.8-425.6	84	1.5-872.5
Intermediate	85	425.7-887.6	84	872.6-1511.2
High	85	887.7-2447.0	84	1511.3-6118.5
Population age <18yrs tertiles (%)
Low	85	2.0-19.7	84	0.0-18.7
Intermediate	85	19.8-26.3	84	18.8-25.5
High	85	26.4-43.4	84	25.6-43.6
ParkIndex values	255	44.6 (33.1)	252	73.0 (29.0)

To illustrate the juxtaposition of ParkIndex values and neighborhood sociodemographic attributes, Figure 1AandB show the distribution of ParkIndex values overlayed with median household income for all block groups in Greenville County and Raleigh, respectively. Table 2 presents differences in ParkIndex scores by block group characteristics across the full sample. Mean ParkIndex values did not differ significantly according to tertiles of median household income (F = 0.15, P = .86), percent non-Hispanic White race/ethnicity (F = 0.22, P = .80), or population density (F = 0.70, P = .50).

Figure 1.

Block Group ParkIndex Scores and Median Household Income Tertiles in (A) Greenville County, SC and (b) Raleigh, NC

Table 2.

Differences in ParkIndex Scores by Block Group Characteristics

Block group Characteristic	N	ParkIndex valueMean (SD)	F (p)^*
	507	58.7 (34.2)	-
Median household income
Low	169	59.6 (31.2)
Intermediate	169	57.8 (35.9)	0.15 (0.86)
High	169	58.7 (35.4)
Percent non-Hispanic White race
Low	169	60.0 (32.7)
Intermediate	169	57.7 (35.0)	0.22 (0.80)
High	169	58.4 (35.0)
Population density
Low	169	57.4 (36.6)
Intermediate	169	61.0 (34.2)	0.70 (0.50)
High	169	57.7 (31.6)

*One way analysis of variance (ANOVA) estimates are adjusted for city location and percentage of population < 18 years.

Discussion

This was the first study, to our knowledge, to examine how a composite metric like ParkIndex differs by income, race/ethnicity, and population density across and within multiple communities. We observed no significant differences in ParkIndex by income, race/ethnicity, and population density in two Southeastern US cities. This is consistent with some previous studies that showed no disparities or clear patterns across different sociodemographic groups in access to walkable built environment attributes such as parks and public open spaces.^28-30 For example, a study conducted in Melbourne, Australia found that there was no significant difference in the availability of public open spaces across different socioeconomic status areas.³⁰ In Kansas City, MO, mixed findings were observed in that lower income census tracts contained more parks, but also had fewer parks with playgrounds and more quality concerns per park.³¹ Our study adds to such findings and extends them by testing a novel spatial park metric (ParkIndex) that incorporates indicators of park quality (as well as access) in relation to socioeconomic and geographic disparities. ParkIndex can offer several unique opportunities to research park equity as it provides a quantitative, multi-dimensional measure of park access, including three empirically driven aspects of parks – number, total space, and quality of parks. As a standardized measure, ParkIndex can also be applied by diverse users without special expertise.

Mitigating the physical activity gap between low and high socioeconomic status areas can be an important step in reducing health disparities, a major public health issue in many countries.³² Inequitable distribution of built environment attributes conducive to physical and mental health, such as parks and open spaces, across low vs high socioeconomic status areas may be one explanation for this gap. However, our findings in these contexts did not support the general hypothesis that economically and socially disadvantaged populations have poorer access to parks and open spaces.^8,33,34 Different measures and tools were used by previous studies to define, conceptualize, and measure access to parks, so increased standardization and application of metrics (such as ParkIndex) would be valuable.

ParkIndex is significant for enhancing public health in that it provides an evidence-based tool to assist citizens and professionals in better conceptualizing environmental characteristics driving park use.^18-20 ParkIndex was developed, in part, to identify and address socioeconomic, race-based, and geographic disparities in park use and park-based PA.^35-39 Perhaps surprisingly, our study did not show clear disparities in ParkIndex values in these two major Southeastern locations. It may be that conscious efforts have been undertaken in these locales to address park equity. Therefore, future research should explore current or historical policies that would affect the present distribution of park resources in these locations and more broadly. It may also be that a composite metric like ParkIndex – which was empirically developed and comprises factors related to both park availability and quality – tends to obfuscate differences in its component parts. Consequently, future research may wish to build on ours by studying whether the various pieces of ParkIndex – number of parks, total park acreage, and average park quality score for all parks within a one half-mile network buffer – are equitably distributed according to geographic or demographic factors. Some research does support this idea, showing that park quality has more marked disparities according to area-level socioeconomic indicators compared to metrics related to park access such as distance to or area of green space.^31,40,41

This study had some limitations. Since our analysis focused on all block groups within only two metropolitan areas, the findings cannot be generalized to all Southeastern US cities or beyond. There was also a minor temporal difference between park data and block group information, although both are likely quite stable. As well, although ParkIndex incorporates comprehensive park availability and quality measures, it cannot truly detect other park attributes, such as the topological aspect of park layouts which may be important.⁴² Moreover, ParkIndex does not include any metrics related to park programming or subjective perceptions such as perceived safety. Lastly, one potentially important park access variable not included in the ParkIndex algorithm was distance (e.g., distance to the closest park), a decision supported by the inconsistent (and often counterintuitive) relationship between distance and park use or physical activity in past research.^10,43,44 Future studies may elect to integrate these attributes into ParkIndex and test whether they differ across different socioeconomic and geographic contexts.

Conclusion

The current study contributes to the fast-growing body of literature examining whether park access is equitably distributed in diverse locations. By testing a unique index of park access, we found no significant differences in ParkIndex values by income, race/ethnicity, and population density in two Southeastern US areas. Future research and practice can investigate the utility and generalizability of this composite metric to identify and mitigate potential environmental injustices and health disparities.

So What?

What Is Already Known on This Topic?

Access to parks is often not equitably distributed within communities, but most studies have not incorporated a comprehensive metric that includes both park availability and diverse quality indicators.

What Does This Article Add?

An innovative tool derived from composite park availability and quality was used to examine whether ParkIndex scores for block groups differed significantly based on income, race/ethnicity, and population density. Although the findings in these settings did not reveal significant results, this methodology could be applied in other contexts to identify park deserts within communities and reveal potential disparities in combined park access and quality according to various socio-economic factors.

What Are the Implications for Health Promotion Practice or Research?

ParkIndex advances public health by offering an evidence-based instrument that helps researchers, professionals, and the general public gain a clearer understanding of access to quality parks within and across communities. ParkIndex can inform decisions and strategies aimed at optimizing park accessibility, ultimately contributing to increased park use and its associated health benefits for individuals and communities.

Footnotes

Acknowledgments

This study was supported by the National Cancer Institute under award number R21CA202693. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

ORCID iDs

Andrew T. Kaczynski

J. Aaron Hipp

Authors’ Contributions

ATK: Conceptualization, Data acquisition, Writing-original draft.

MEW: Conceptualization, Data acquisition, Writing-original draft, Statistical analysis.

JS: Conceptualization, Data acquisition, Writing-review and editing, Statistical analysis.

SMH: Conceptualization, Data acquisition, Writing-review and editing.

EWS: Conceptualization, Data acquisition, Writing-review and editing.

JAH: Conceptualization, Data acquisition, Writing-review and editing.

FHS: Writing-original draft.

MJK: Writing-original draft.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the National Cancer Institute under award number R21CA202693. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.*

References

Kaczynski

Henderson

. Environmental correlates of physical activity: a review of evidence about parks and recreation. Leis Sci. 2007;29(4):315-354. doi:10.1080/01490400701394865

Koohsari

Mavoa

Villanueva

, et al. Public open space, physical activity, urban design and public health: concepts, methods and research agenda. Health Place. 2015;33:75-82. doi:10.1016/j.healthplace.2015.02.009

Reyes-Riveros

Altamirano

De La Barrera

Rozas-Vásquez

Vieli

Meli

. Linking public urban green spaces and human well-being: a systematic review. Urban For Urban Green. 2021;61:127105. doi:10.1016/j.ufug.2021.127105

Akpinar

. Urban green spaces for children: a cross-sectional study of associations with distance, physical activity, screen time, general health, and overweight. Urban For Urban Green. 2017;25:66-73. doi:10.1016/j.ufug.2017.05.006

Besenyi

Kaczynski

Stanis

SAW

Bergstrom

Lightner

Hipp

. Planning for health: a community-based spatial analysis of park availability and chronic disease across the lifespan. Health Place. 2014;27:102-105. doi:10.1016/j.healthplace.2014.02.005

Schipperijn

Cerin

Adams

, et al. Access to parks and physical activity: an eight country comparison. Urban For Urban Green. 2017;27:253-263. doi:10.1016/j.ufug.2017.08.010

Bruton

Floyd

. Disparities in built and natural features of urban parks: comparisons by neighborhood level race/ethnicity and income. J Urban Health. 2014;91(5):894-907. doi:10.1007/s11524-014-9893-4

Hughey

Walsemann

Child

Powers

Reed

Kaczynski

. Using an environmental justice approach to examine the relationships between park availability and quality indicators, neighborhood disadvantage, and racial/ethnic composition. Landsc Urban Plann. 2016;148:159-169. doi:10.1016/j.landurbplan.2015.12.016

Jones

Moore

, et al. Disparities in physical activity resource availability in six US regions. Prev Med. 2015;78:17-22. doi:10.1016/j.ypmed.2015.05.028

10.

Kaczynski

Besenyi

Stanis

SAW

, et al. Are park proximity and park features related to park use and park-based physical activity among adults? Variations by multiple socio-demographic characteristics. Int J Behav Nutr Phys Act. 2014;11(1):146. doi:10.1186/s12966-014-0146-4

11.

Mavoa

Koohsari

Badland

, et al. Area-level disparities of public open space: a geographic information systems analysis in metropolitan Melbourne. Urban Pol Res. 2015;33(3):306-323. doi:10.1080/08111146.2014.974747

12.

Rigolon

Browning

Jennings

. Inequities in the quality of urban park systems: an environmental justice investigation of cities in the United States. Landsc Urban Plann. 2018;178:156-169. doi:10.1016/j.landurbplan.2018.05.026

13.

Higgs

Fry

Langford

. Investigating the implications of using Alternative GIS-based Techniques to measure accessibility to green space. Environ Plann B Plann Des. 2012;39(2):326-343. doi:10.1068/b37130

14.

James

Berrigan

Hart

, et al. Effects of buffer size and shape on associations between the built environment and energy balance. Health Place. 2014;27:162-170. doi:10.1016/j.healthplace.2014.02.003

15.

Lachowycz

Jones

. Greenspace and obesity: a systematic review of the evidence. Obes Rev. 2011;12(5):e183-e189. doi:10.1111/j.1467-789X.2010.00827.x

16.

Sugiyama

Francis

Middleton

Owen

Giles-Corti

. Associations between recreational walking and attractiveness, size, and proximity of neighborhood open spaces. Am J Public Health. 2010;100(9):1752-1757. doi:10.2105/AJPH.2009.182006

17.

Talen

Anselin

. Assessing spatial equity: an evaluation of measures of accessibility to public playgrounds. Environ Plan A. 1998;30(4):595-613. doi:10.1068/a300595

18.

Kaczynski

Schipperijn

Hipp

, et al. ParkIndex: development of a standardized metric of park access for research and planning. Prev Med. 2016;87:110-114. doi:10.1016/j.ypmed.2016.02.012

19.

Kaczynski

Hughey

Stowe

, et al. ParkIndex: validation and application of a pragmatic measure of park access and use. Prev Med Rep. 2020;20:101218. doi:10.1016/j.pmedr.2020.101218

20.

Oliphant

Hughey

Stowe

Kaczynski

Schipperijn

Hipp

. ParkIndex: using key informant interviews to inform the development of a new park access evaluation tool. JPRA 2019;37(1). doi:10.18666/JPRA-2019-8926

21.

Hughey

Wende

Stowe

Kaczynski

Schipperijn

Hipp

. Frequency of neighborhood park use is associated with physical activity among adults in four US cities. J Phys Act Health. 2021;18(5):603-609. doi:10.1123/jpah.2020-0540

22.

Cutts

Darby

Boone

Brewis

. City structure, obesity, and environmental justice: an integrated analysis of physical and social barriers to walkable streets and park access. Soc Sci Med. 2009;69(9):1314-1322. doi:10.1016/j.socscimed.2009.08.020

23.

Floyd

Johnson

. Coming to terms with environmental justice in outdoor recreation: a conceptual discussion with research implications. Leisure Sci. 2002;24(1):59-77. doi:10.1080/01490400252772836

24.

Taylor

Floyd

Whitt-Glover

Brooks

. Environmental justice: a framework for collaboration between the public health and parks and recreation fields to study disparities in physical activity. J Phys Act Health. 2007;4(s1):S50-S63. doi:10.1123/jpah.4.s1.s50

25.

United States Census Bureau . American community survey data. The United States census Bureau. https://www.census.gov/programs-surveys/acs/data.html. Accessed 5 November 2020.

26.

Besenyi

Diehl

Schooley

, et al. Development and testing of mobile technology for community park improvements: validity and reliability of the eCPAT application with youth. Transl Behav Med. 2016;6(4):519-532. doi:10.1007/s13142-016-0405-9

27.

Kaczynski

Wilhelm Stanis

Besenyi

. Development and testing of a community stakeholder park Audit tool. Am J Prev Med. 2012;42(3):242-249. doi:10.1016/j.amepre.2011.10.018

28.

Lotfi

Koohsari

. Measuring objective accessibility to neighborhood facilities in the city (A case study: zone 6 in Tehran, Iran). Cities. 2009;26(3):133-140. doi:10.1016/j.cities.2009.02.006

29.

Pearce

Witten

Hiscock

Blakely

. Are socially disadvantaged neighbourhoods deprived of health-related community resources? Int J Epidemiol. 2007;36(2):348-355. doi:10.1093/ije/dyl267

30.

Timperio

Ball

Salmon

Roberts

Crawford

. Is availability of public open space equitable across areas? Health Place. 2007;13(2):335-340. doi:10.1016/j.healthplace.2006.02.003

31.

Vaughan

Kaczynski

Wilhelm Stanis

Besenyi

Bergstrom

Heinrich

. Exploring the distribution of park availability, features, and quality across Kansas city, Missouri by income and race/ethnicity: an environmental justice investigation. Ann Behav Med. 2013;45(S1):28-38. doi:10.1007/s12160-012-9425-y

32.

Frieden

. Strategies for Reducing Health Disparities—Selected CDC-Sponsored Interventions, United States, 2014; 2014.

33.

King

Clarke

. A disadvantaged advantage in walkability: findings from socioeconomic and geographical analysis of national built environment data in the United States. Am J Epidemiol. 2015;181(1):17-25. doi:10.1093/aje/kwu310

34.

Wüstemann

Kalisch

Kolbe

. Access to urban green space and environmental inequalities in Germany. Landsc Urban Plann. 2017;164:124-131. doi:10.1016/j.landurbplan.2017.04.002

35.

Baran

Smith

Moore

, et al. Park use among youth and adults: examination of individual, social, and urban form factors. Environ Behav. 2014;46(6):768-800. doi:10.1177/0013916512470134

36.

Carlson

Brooks

Brown

Buchner

. Racial/Ethnic differences in perceived access, environmental barriers to use, and use of community parks. Prev Chronic Dis. 2010;7(3):A49. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2879981/. Accessed January 15, 2020.

37.

Cohen

Marsh

Williamson

, et al.

Parks and physical activity: why are some parks used more than others?

Prev Med. 2010;50:S9-S12. doi:10.1016/j.ypmed.2009.08.020

38.

Cohen

Han

Park

Williamson

Derose

. Park use and park-based physical activity in low-income neighborhoods. J Aging Phys Act. 2018;27:1-9. doi:10.1123/japa.2018-0032

39.

Floyd

Spengler

Maddock

Gobster

Suau

. Park-based physical activity in diverse communities of two U.S. Cities. Am J Prev Med. 2008;34(4):299-305. doi:10.1016/j.amepre.2008.01.009

40.

Engelberg

Conway

Geremia

, et al. Socioeconomic and race/ethnic disparities in observed park quality. BMC Public Health. 2016;16(1):395. doi:10.1186/s12889-016-3055-4

41.

Hobbs

Green

Griffiths

, et al. Access and quality of parks and associations with obesity: a cross-sectional study. SSM Popul Health. 2017;3:722-729. doi:10.1016/j.ssmph.2017.07.007

42.

Zhai

Baran

. Do configurational attributes matter in context of urban parks? Park pathway configurational attributes and senior walking. Landsc Urban Plann. 2016;148:188-202. doi:10.1016/j.landurbplan.2015.12.010

43.

Huang

Guo

. How do travel distance and park size influence urban park visits? Urban For Urban Green. 2020;52:126689. doi:10.1016/j.ufug.2020.126689

44.

Wang

Brown

Liu

Mateo-Babiano

. A comparison of perceived and geographic access to predict urban park use. Cities. 2015;42:85-96. doi:10.1016/j.cities.2014.10.003