Health Equity in the Built Environment: Piloting a Walkability Index in San Francisco

doi:10.21203/rs.3.rs-2753715/v1

Download PDF

Research Article

Health Equity in the Built Environment: Piloting a Walkability Index in San Francisco

https://doi.org/10.21203/rs.3.rs-2753715/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The purpose of this study is to develop and test a local measure of walkability in San Francisco County and evaluate walkability by race, ethnicity, and educational attainment at the census block group level. We used data collected as part of the 2019–2020 Bay Area Young Adult Health Study, wherein we conducted audits of 1525 randomly selected census blocks to assess domains, such as aesthetics, access to amenities, street safety, and topography, which we then imputed to the remaining blocks in the city for overall coverage. We linked the walkability index to 2015–2019 American Community Survey Data aggregating walkability to the block group level. We performed a geographically weighted regression to measure associations between walkability, race/ethnicity, and educational attainment and found significant negative associations between walkability and the percentage of non-Hispanic Black residents in specific neighborhoods and non-Hispanic Asian residents in all neighborhoods. We also found significant positive associations between the percentage of residents with a bachelor’s or higher degree and walkability. Inequities in walkability across neighborhoods in San Francisco County largely mirrored sociodemographic inequities. Our approach to constructing a walkability index points to a relatively efficient means of estimating fine-scale neighborhood characteristics in urban areas, and our findings suggest that investments in neighborhood infrastructure may positively impact walkability and address broader inequity concerns.

Walkability measures and their associated outcomes have proliferated in the public health and planning literatures over the past twenty years.¹ Walking is one of the most common forms of physical activity and studies have consistently linked walking to many positive health outcomes, including lower risks of obesity,^{2, 3} diabetes,⁴ cardiovascular disease,⁵ cognitive impairment,⁶ and mental illness.⁷ However, the walkability of an area is dependent on its built environment, or its human-constructed landscape, which includes both obstacles that decrease (e.g., safety concerns, traffic volume, sidewalk maintenance, highway interdiction) and resources that improve walkability (e.g., accessibility to parks and other amenities, social interaction).^8–11 These obstacles and resources in turn depend on broader policy decisions, such as zoning laws restricting or promoting healthful urban planning,¹² and systemic processes, such as persistent disinvestment that may discourage local business development.¹³ Additionally, an area’s topography has been shown to influence perceptions of walkability among urban residents.¹⁴

Myriad studies have sought to capture walkability using a variety of measures. Composite indices of observed environmental features, single item measures, self-reported measures, and spatially constructed measures have been tested and applied in various settings.^{8, 10, 11, 15–19} More recent studies have therefore focused on summarizing and evaluating walkability measurements for the purpose of providing a generalizable approach to estimating walkability across diverse geographic regions.^{1, 20} This is an appropriate focus when seeking to draw inferences about walkability at broad national or regional levels and to meet the scientific goals of reproducibility and generalizability. However, we argue that walkability is fundamentally a local consideration and that understanding the particular contexts of one urban area or set of neighborhoods at a time has more utility for local and regional policymakers who have responsibility for urban planning decisions than a broader application of walkability meant to capture key features of urban areas in general.

This is not to suggest that the existing walkability literature, which is extensive and highly informative, is seeking the wrong target; rather, we argue that the broader literature is essential for clarifying population-level health implications related to walkability, but there is also a need to conduct local area studies that provide local policymakers with intervention targets. Additionally, when studies focus on sets of neighborhoods for which other information is available, such as individual behavioral data, we can gain insight not only into the characteristics of the built environment generally, but also into how specific populations interact with their own neighborhood environments.

To this end, we conducted a pilot study of neighborhood walkability in San Francisco County in which data were collected in a randomly selected set of census blocks (“neighborhoods”) from the universe of census blocks in which respondents to our corresponding probabilistic household survey of young adults resided (N = 1363). San Francisco is an ideal area for this study in some ways as the City and County government are the same entity, providing one policy target, and the county is only 7x7 square miles. Nevertheless, the county suffers from the same residential segregation²¹ as most cities exhibit across the United States and has substantial topographic diversity.

The first objective of this study is to develop a walkability index for San Francisco City and County (SFC) using area-level data collected in SFC, and in doing so, to test the utility of multiply imputing neighborhood characteristics from randomly selected census blocks to the city as a whole. The second objective is to determine how well our walkability index reflects longstanding inequities in neighborhood investment between San Francisco neighborhoods. With respect to the latter objective, we hypothesize that neighborhoods with greater proportions of residents racialized as Black, Asian and Latino, compared to those racialized as White, and neighborhoods with lower educational attainment than the county average, will also be less walkable neighborhoods.

Sample Characteristics

The 2014–2020 Bay Area Young Adult Health Panel (BAYAHS) study was a probabilistic longitudinal survey of young adults (18–26) in Alameda and San Francisco Counties in California.²² As part of this study, we also collected census block-level observational data in a subset of participant neighborhoods to capture contextual characteristics of the areas in which young adults resided. From 1525 (of 31210 in the two counties) census blocks where at least one participant resided, we used Stata v16 to randomly select 25% of blocks in Fremont, Oakland, and San Francisco, where the largest young adult concentrations were, and 15% of blocks in all other areas of the two counties, which had lower spatial concentrations of young adults. Pairs of research assistants, all of whom resided in the two counties, were racially and ethnically diverse, and spoke English, Spanish, and Cantonese, were sent to each selected block to walk the entirety of the block. They would then complete an assessment instrument, programmed in ArcGIS Survey123, recording various features of the block, including aesthetics, access to amenities, sidewalk safety and maintenance, and other factors. Each block therefore had two assessments, and we conducted reliability tests to ensure general inter-rater agreement between the two for each block. We then randomly selected one report for each block to retain for analysis of walkability. For this study, we chose to focus on San Francisco County (N = 138 of 7348 blocks) because the topographical and architectural environment differs substantially from that in Alameda County, as does the scale of the county. We linked the neighborhood data to 2015–2019 American Community Survey data, accessed via the National Historical GIS website,²³ at the block group level to account for area-level differences in sociodemographic characteristics. This study was exempt from IRB review.

Measures

Dependent Variables

The dependent variable for our analysis is a walkability index based on a composite of seven measures from the 2019-20 BAYAHS neighborhood assessment instrument, six of which had 5-point Likert response scales: “How attractive are the buildings/homes? (Not at all to extremely attractive);” “How much green or open space do you see (such as parks, parklets, cemeteries, sports fields?) (None to a great deal);” “How many amenities are there nearby (such as coffee shops, dry cleaning services, beauty shops)? (None to a great many);” “How many vacant lots do you see? (None to a great many);” “Indicate the level of sidewalk safety (i.e., there are ample natural barriers, spaces, or parked cars to separate pedestrians from traffic) (Terrible to excellent);” “Are there streetlights? (None to a great many).” The final measure had a 7-point Likert response and was included specifically to account for San Francisco’s undulating topography: “How hilly are the streets? (Extremely hilly to extremely flat).” The indexed measure was aggregated to the block group level as educational attainment and annual income (the latter having been tested but excluded from final analyses as an irrelevant variable) are not available at the census block level.

Independent variables

Race was measured as (1) the percentage of the population in the census block group that was non-Hispanic Black, and (2) the percentage of the population that was non-Hispanic Asian, with the remaining race groups (primarily non-Hispanic White) as the reference category. Ethnicity was measured as the percentage of the population in the census block group that was Latino. Educational attainment was used to approximate socioeconomic status, measured as the percentage of the population with a bachelor’s degree or higher level of education. Income was included in earlier models but had high multicollinearity with education and little effect on results.

Analysis

The variables used to construct the walkability index were selected in a two-step process. First, we consulted the literature to determine what measures of walkability were the most theoretically significant. This yielded three thematic areas: Neighborhood aesthetics, land use, and safety.^{8, 19, 24} We therefore selected all variables corresponding to these themes in the BAYAHS instrument and conducted separate principal components analysis for the variables in each thematic area using the stats and factoextra package in R (version 1.4.1106). We retained all components with an eigenvalue > 1, which resulted in six of the seven measures used to construct the walkability index. Finally, we included “How hilly are the streets?” to adjust for San Francisco’s hilly landscape. As the last measure used a 7-point Likert response scale and the rest used a 5-point Likert response scale, we standardized all index variables to ensure comparability. We then summed the values of the seven measures, with an equal weight assigned to each. Because the scores were standardized, they range from negative to positive values. Negative values were in areas with walkability scores lower than the mean, and positive values were in areas with scores above mean walkability.

To assign walkability values to the census blocks that were not included in the sample, we imputed scores using the mice package (multiple imputation using chained equations)²⁵ in R, controlling for population density (in square miles), percent of the population that was non-Hispanic White, non-Hispanic Black, non-Hispanic Asian, or Latino, percent of homes that were occupied, and the dependency ratio (population 18–64/population under the age of 18 + population over 65). The imputed values were then visualized in ArcGIS Pro v2.8 and aggregated to the block group level. We conducted a Moran’s I test on the walkability measure with the spdep package in R, which returned a p-value < 0.001, indicating there was significant spatial autocorrelation in walkability across the county. We therefore used MGWR 2.2 to perform Gaussian geographically weighted regression of walkability on the ACS block group level variables with an adaptive kernel.

Table 1 shows the mean walkability score for SFC and block group population characteristics. Non-Hispanic White residents represent the largest racial group in SFC, followed by non-Hispanic Asian residents, Latino residents, and non-Hispanic Black residents. Just over half the population had earned a bachelor's or higher degree, substantially higher than the national average of 33.13%.²⁶

Table 1

*Descriptive Statistics* (N = 578 census block groups)
Variable	Mean	Std Dev
Outcome
Walkability (standardized)	-20.36	1.55
Race, Ethnicity & Education
%Non-Hispanic White (referent)	42.89%	22.29
%Non-Hispanic Black	4.68%	7.99
%Non-Hispanic Asian	32.96%	20.28
%Latino	14.06%	12.64
%Earned Bachelor's Degree or Higher	58.76%	21.59

Table 2 shows the global GWR results. Block groups with greater percentages of non-Hispanic Black and Asian residents, compared to non-Hispanic White, had walkability scores lower than the county average. Block groups with greater percentages of residents with a bachelor's or higher degree had walkability scores higher than the county averages. Specifically, for every reduction of 44% and 51% in the composition of non-Hispanic Black and Asian residents, respectively, walkability scores increased by 1, and for every 20% increase in residents with higher educational attainment, walkability scores increased by 1. There was no significant relationship between percentage of Latino residents and walkability score.

Table 2

*Global MGWR Results* (n = 578 census block groups)
Variable	β	SE	Significance level
Intercept	0.000	0.026
%Non-Hispanic White (referent)	-	-
%Non-Hispanic Black	-0.438	0.032	***
%Non-Hispanic Asian	-0.512	0.043	***
%Latino	0.052	0.037
%Earned Bachelor's Degree or higher	0.195	0.049	***
R² = 0.168 *** p < 0.001

Figure 1 illustrates the highest and lowest walkability scores in the county. The areas with the highest walkability scores, illustrated in purple, are concentrated in the central and northern San Francisco neighborhoods, including Haight-Ashbury, the Castro, the Mission, Noe Valley, Pacific Heights and the Marina, North Beach, the Financial District, and the Tenderloin/Nob Hill. The areas with the lowest walkability scores, illustrated in green, are primarily in the Bayview Hunters Point, Visitacion Valley, Excelsior and Ocean View neighborhoods. The Western Addition also had low walkability scores but was surrounded by areas with high scores.

Figure 1 also demonstrates examples of how well our imputed walkability index performed against the commonly used Walk Score®, which is based on access and distance to amenities, population density, and road metrics.²⁷ Walk Score lists San Francisco as one of the top three most walkable cities in the country, so even in areas with lower scores, they remain above 50. However, the Walk Scores in the areas with lower walkability according to our index have much lower scores than those in areas of high walkability, and while Chinatown and downtown San Francisco rank as the 1st and 2nd most walkable neighborhoods according to Walk Score, Oceanview ranks 76th and Bayview Hunters Point ranks 84th .²⁷

The neighborhoods with the lowest walkability scores, on average, were also those where there was an inverse relationship between the percentage of non-Hispanic Black residents and walkability scores, as shown in Fig. 2. The largest associations between lower walkability scores and the percentage of non-Hispanic Black residents were in the Western Addition, the Castro and Noe Valley, Bayview Hunters Point, and Ocean View. Alternatively, in Russian Hill there was a positive association between walkability scores and the percentage of non-Hispanic Black residents. Interestingly, the percentage of non-Hispanic Asian residents per census block group was negatively associated with walkability scores for the entire county (not shown). However, the parameter coefficients for non-Hispanic Asian residents’ walkability ranged from − 0.22 to -0.24, indicating a lesser effect of the association than between non-Hispanic Black residents and low walkability scores.

Higher educational attainment was positively associated with walkability everywhere except the Outer Sunset, Parkside, Twin Peaks, Outer Mission, Mission and South of Market neighborhoods where there was no association, as shown in Fig. 3. The strongest effects were in the central part of the county, from Bayview Hunters Point in the south to the Marina in the north as well as in the Outer Richmond in the northwest.

The first objective of this study was to test the performance of our localized walkability measure for which we calculated values from data we collected at the census block level and imputed values to the remaining census blocks in San Francisco City-County. Our walkability index closely corresponded to neighborhood Walk Scores® and the results make logical sense given the cultural and physical topographies of SFC. Thus, designing a stratified or simple random sample of neighborhood geographies, depending on study objective or outcome of interest, and imputing the values from the sample geographies to the larger region may be an efficient and accurate way to evaluate local environments for health behavior. We sampled census blocks in our design to obtain observational data at a very fine scale, which provides a level of detail to build from that data collection in larger administrative or observed boundaries may not. Such local specification may ensure more accuracy in imputed results as well as more specific local targets for policy and behavioral interventions.

We also hypothesized in this study first, that neighborhoods with larger proportions of residents racialized as Black, Asian and Latino, and second, neighborhoods with lower educational attainment than average, would also have significantly lower walkability scores. Regarding our first hypothesis, we did find significant inverse relationships between walkability and the percentage of non-Hispanic Black and non-Hispanic Asian residents, but not for Latino residents. Our second hypothesis proved apt for most neighborhoods; educational attainment was positively associated with walkability across the county, except in the Mission/Potrero Hill, the Outer Mission, Outer Sunset, and Parkside where there was no significant relationship.

In terms of overall walkability, most of the census block groups were within one standard deviation of the mean value for our dataset. Additionally, a large majority of the census block groups had walkability values greater than the median value for our dataset. These results speak to San Francisco’s status as one of the nation’s most walkable cities.²⁸ However, despite generally high levels of walkability, there are still notable sociodemographic inequities between neighborhoods.

There are very clear differences in sociodemographic characteristics between the areas with greater walkability, average walkability, and lower than average walkability (shown in purple, white and green in Fig. 1, respectively), which likely explain our results in part. Supplemental Table 1 includes 2015–2019 ACS estimates of these characteristics, showing that the percentage of non-Hispanic White residents in the average walkability areas was 160% higher than in the less walkable neighborhoods, and 75% higher in the most walkable neighborhoods. Alternatively, the non-Hispanic Black population concentration drops 81% between the low and average walkability neighborhoods and drops a further 12% from the average to high walkability neighborhoods. Interestingly, the low and average walkability neighborhoods share a similar proportion of non-Hispanic Asian residents (44.7% low, 43.4% average), while the high walkability areas have a much lower percentage of non-Hispanic Asian residents (18.5%), which partially explains the significant inverse association between walkability and percentage of non-Hispanic Asian residents across the county. Median income in the lowest walkability neighborhoods was $74,880 compared to $113,572 in the average walkability neighborhoods and $148,776 in the most walkable neighborhoods. The percentages of residents below 200% of poverty were 33%, 21% and 16% respectively.

Bayview Hunters Point and Western Addition, two of the neighborhoods with lower than average walkability were also neighborhoods with the highest proportions of non-Hispanic Black residents, which have also been neighborhoods historically designated as “Hazardous” or “Definitely Declining” in the era of federal redlining, preventing issuance of government-backed home loans in the neighborhood.²⁹ Although redlining is technically illegal today, disinvestment in these communities has continued across decades, which impacts features like walkability through a lack of government funding and resources to support new business development, green space, street maintenance, and general neighborhood upkeep.³⁰ Additionally, the Bayview Hunters Point neighborhood is a heavily industrialized “sacrifice zone,”³¹ with an abandoned naval shipyard designated as a Superfund site, more than 150 brownfields, the former Candlestick Park stadium, which has been torn down but not remediated, and a sewage treatment plant that treats more than 80% of San Francisco’s sewage.³² These factors reduce aesthetic value for walking and expose residents to excessive risk of air pollution while also reducing home and property values, keeping people in place.³²

Non-Hispanic Asian residents have not been as segregated as Non-Hispanic Black residents in San Francisco; however, between 2000 and 2020 they lost ground in the more walkable neighborhoods, as shown in Supplemental Figs. 1a and 1b, expanded further into less walkable neighborhoods, and have historically resided in a semicircle around the outsides of the most walkable neighborhoods that cluster in the city center. The infusion of technology funding and salaries over the past two decades, the increase in short-term rental properties, foreign investment in local property, and failures to expand affordable housing have priced many communities out of the neighborhoods in which they have resided for generations.^33–35 This has pushed long-term and intergenerational residents to less desirable neighborhoods with fewer amenities, less aesthetic appeal, and greater exposure to hazards. Our walkability index provides an initial indicator of these inequities while also providing a more direct assessment of health promoting behavior. Engaging in urban planning activities designed to improve walkability may have ripple effects that increase overall neighborhood wellness. The point at which this becomes a trigger for gentrification, however, is a concern and remains to be seen.

There are a few limitations to our study. First, our walkability index was derived from a local observational study of neighborhoods in the Bay Area, an area with distinct topography, architecture, and population composition, and the results may not be generalizable to other cities in the US. However, we contend that the method certainly could be applied in other cities and towns where researchers, public health professionals, and policymakers are interested in collecting local data how their own built environments may be improved for health and wellness. A second limitation was that our research assistants who collected the neighborhood data, though they lived in the study area, did not reside in most of the surveyed blocks, and they completed neighborhood audits based on their own perceptions of the census blocks they surveyed. This may introduce bias as it cannot accurately capture the residents’ perceptions of walkability for their own neighborhoods. Nevertheless, the method we employed is a relatively efficient and affordable way to measure neighborhoods for health behavior across an urban area, and in this case, correlated well with sociodemographic inequities in San Francisco.

Health behaviors, like walking, are fundamentally local activities that require investment from local policy actors. We contend that while more general walkability measures are appropriate for conducting comparative analyses across urban areas, local measures may be more instructive for policy and behavioral intervention strategies. Additionally, in San Francisco, as has been documented in many places, walkability scores tend to correspond with sociodemographic inequities and spatial segregation. Targeted community-based, planning, and policy efforts to address inequities broadly, or walkability specifically, are both likely to improve overall health and mitigate health disparities.

Dalmat RR, Mooney SJ, Hurvitz PM, Zhou C, Moudon AV, Saelens BE. Walkability measures to predict the likelihood of walking in a place: A classification and regression tree analysis. Health & Place. 2021/11/01/ 2021;72:102700.
Brown BB, Yamada I, Smith KR, Zick CD, Kowaleski-Jones L, Fan JX. Mixed land use and walkability: Variations in land use measures and relationships with BMI, overweight, and obesity. Health & Place. 2009;15(4):1130-1141.
Lovasi GS, Neckerman KM, Quinn JW, Weiss CC, Rundle A. Effect of Individual or Neighborhood Disadvantage on the Association Between Neighborhood Walkability and Body Mass Index. American Journal of Public Health. February 1, 2009 2009;99(2):279-284.
Creatore MI, Glazier RH, Moineddin R, et al. Association of neighborhood walkability with change in overweight, obesity, and diabetes. Jama. 2016;315(20):2211-2220.
Howell NA, Tu JV, Moineddin R, Chu A, Booth GL. Association between neighborhood walkability and predicted 10‐year cardiovascular disease risk: The CANHEART (Cardiovascular Health in Ambulatory Care Research Team) Cohort. Journal of the American Heart Association. 2019;8(21):e013146.
Sylvers DL, Hicken M, Esposito M, Manly J, Judd S, Clarke P. Walkable neighborhoods and cognition: implications for the design of health promoting communities. Journal of aging and health. 2022:08982643221075509.
James P, Hart JE, Banay RF, Laden F, Signorello LB. Built environment and depression in low-income African Americans and Whites. American Journal of Preventive Medicine. 2017;52(1):74-84.
Ewing R, Handy S. Measuring the Unmeasurable: Urban Design Qualities Related to Walkability. Journal of Urban Design. 2009/02/01 2009;14(1):65-84.
Fonseca F, Ribeiro PJG, Conticelli E, et al. Built environment attributes and their influence on walkability. International Journal of Sustainable Transportation. 2021:1-40.
Frank LD, Sallis JF, Saelens BE, et al. The development of a walkability index: application to the Neighborhood Quality of Life Study. British Journal of Sports Medicine. 2010;44(13):924-933.
Deng C, Dong X, Wang H, et al. A Data-Driven Framework for Walkability Measurement with Open Data: A Case Study of Triple Cities, New York. ISPRS International Journal of Geo-Information. 2020;9(1):36.
Taylor D. The Rise of Racial Zoning: Residential Segregation. Toxic Communities: Environmental Racism, Industrial Pollution, and Residential Mobility. New York, NY: NYU Press; 2014.
Taylor D. Toxic Communities: Environmental Racism, Industrial Pollution, and Residential Mobility: NYU Press; 2014.
Leslie E, Saelens B, Frank L, et al. Residents’ perceptions of walkability attributes in objectively different neighbourhoods: a pilot study. Health & Place. 2005/09/01/ 2005;11(3):227-236.
Giles-Corti B, Broomhall MH, Knuiman M, et al. Increasing Walking: How Important is Distance To, Attractiveness, and Size of Public Open Space? American Journal of Preventive Medicine. 2005;28(2, Supplement 2):169-176.
Gullón P, Badland H, Alfayate S, et al. Assessing Walking and Cycling Environments in the Streets of Madrid: Comparing On-Field and Virtual Audits. Journal of Urban Health. 2015/09/08 2015:1-17.
Leslie E, Coffee N, Frank L, Owen N, Bauman A, Hugo G. Walkability of Local Communities: Using Geographic Information Systems to Objectively Assess Relevant Environmental Attributes. Health & Place. 2007;13(1):111-122.
Lovasi G, Moudon A, Pearson A, et al. Using Built Environment Characteristics to Predict Walking for Exercise. International Journal of Health Geographics. 2008;7(1):10.
Saelens BE, Sallis JF, Frank LD. Environmental correlates of walking and cycling: Findings from the transportation, urban design, and planning literatures. Annals of Behavioral Medicine. April 01 2003;25(2):80-91.
Yi L, Wilson JP, Mason TB, Habre R, Wang S, Dunton GF. Methodologies for assessing contextual exposure to the built environment in physical activity studies: A systematic review. Health & Place. 2019/11/01/ 2019;60:102226.
Landrine H, Corral I. Separate and unequal: residential segregation and black health disparities. Ethnicity & Disease. 2009;19(2):179.
Holmes LM, Popova L, Ling PM. State of transition: Marijuana use among young adults in the San Francisco Bay Area. Preventive Medicine. 9// 2016;90:11-16.
Manson S, Schroeder J, Van Riper D, Kugler T, Ruggles S. IPUMS National Historical Geographic Information System: Version 16.0. In: Center UoMP, ed. Minneapolis, MN; 2021.
Cerin E, Leslie E, du Toit L, Owen N, Frank LD. Destinations that matter: Associations with walking for transport. Health & Place. 2007;13(3):713-724.
Sterne JAC, White IR, Carlin JB, et al. Multiple Imputation for Missing Data in Epidemiological and Clinical Research: Potential and Pitfalls. BMJ. 2009-06-29 12:25:03 2009;338(b2393).
Holmes LM. Author's Analysis of the American Community Survey Point Estimates. Washington D.C.: U.S. Census Bureau; 2019.
Redfin Realty. Walk Score. Seattle, WA 2023.
McMahon S. The 10 Most Walkable Cities in the U.S. Smarter Travel. Boston, MA: Smarter Travel Media LLC; 2021.
Connolly N, Winling L, Nelson RK, Marciano R. Mapping inequality:‘Big data’meets social history in the story of redlining. The Routledge Companion to Spatial History: Routledge; 2018:502-524.
Chapple K, Elias RR. Analyzing investment flows in comprehensive community revitalization: The case of Bayview Hunters Point, San Francisco. Journal of Urban Affairs. 2018;40(4):494-517.
Lerner S. Sacrifice zones: the front lines of toxic chemical exposure in the United States. Cambridge, MA: MIT Press; 2012.
Greenaction for Health & Environmental Justice. Pollution Problems in Bayview Hunters Point. San Francisco, CA 2016.
Sassen S. The Global City: Enabling Economic Intermediation and Bearing Its Costs. City & Community. 2016;15(2):97-108.
Nieuwland S, van Melik R. Regulating Airbnb: how cities deal with perceived negative externalities of short-term rentals. Current Issues in Tourism. 2020/04/02 2020;23(7):811-825.
Stehlin J. The Post-Industrial “Shop Floor”: Emerging Forms of Gentrification in San Francisco's Innovation Economy. Antipode. 2016;48(2):474-493.

Supplement.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Health Equity in the Built Environment: Piloting a Walkability Index in San Francisco

Status:

Version 1

Abstract

Figures

Introduction

Methods

Sample Characteristics

Measures

Dependent Variables

Independent variables

Analysis

Results

Discussion

Conclusion

References

Supplementary Files

Status:

Version 1