THE SOCIAL DETERMINANTS OF VULNERABILITY FRAMEWORK: INCORPORATING THE NEEDS OF PEOPLE INTO MITIGATION, RESPONSE, AND
RECOVERY
A thesis presented by
S. Atyia Martin
To
Doctor of Law and Policy Program
In partial fulfillment of the requirements for the degree of Doctor of Law and Policy
College of Professional Studies
Northeastern University
Boston, Massachusetts
June, 2014
2
DEDICATION
This research project is dedicated to my family. My husband, Roy Martin, has been there
to take care of the homefront while I have worked on this research project. I am grateful for his
love and support in the darkest hours and in the joyous moment of completion. This research
project and my participation in the Northeastern University Doctor of Law and Policy program
would not be possible without him.
My children, Sharoya, Roy Jr., Raekwon, Ryan, and Sonja, have had to deal with my
absence in their lives during this process. I hope they are able to understand the benefits of the
sacrifices that everyone had to make. Finally, my mother, Gloria Walker, helped to take care of
the children and ensured they were able participate in all of the fun things that I was not able to
experience with them.
3
ACKNOWLEDGEMENTS
This project required many different types of qualitative and quantitative analyses. My
advisor, Dr. Neenah Estrella-Luna, coached me through every step of this research with patience
and grace. She spent many late nights reviewing and responding to my drafts and questions. Dr.
Estrella-Luna brought incredible insight and guidance. I am forever grateful for her commitment
to me and the other students she provided the same level of quality support.
Harold Cox, MSSW, was my second reader. He very graciously accepted the role despite
his very busy schedule as the associate dean at the Boston University School of Public Health
and his many board positions advancing public health in the City of Boston and the
Commonwealth of Massachusetts. It has been an honor and a pleasure to work with him.
The staff of the Office of Public Heatlh Preparedness at the Boston Public Heatlh
Commission does such amazing work. I have been able to rely on them to accomplish major
feats, especially after the Boston bombings and the many emergencies that ensued while I
participated in this program. I never missed an intensive seminar because they are so committed
and talented. Finally, I would like to thank Dr. Barbara Ferrer, Commission of Public Health for
the City of Boston, and James Hooley, Chief of Boston Emergency Medical Services, for their
support.
4
ABSTRACT
The social circumstances of people significantly determine the severity of poor outcomes after
disasters. The majority of Americans live in cities that face higher risk because of the density of
infrastructure, assets, and people, particularly vulnerable populations. I developed the Social
Determinants of Vulnerability Framework based on link analysis of the literature to help
planners in cities better identify and understand the most vulnerable people in their area. The
Framework also provides a way to reduce the likelihood of civil rights violations and poor
outcomes for people with limited ability to prepare for, adapt to, and cope with emergencies. The
Framework identifies seven interrelated social factors that seem to be driving vulnerability:
children, people with disabilities, older adults, chronic and acute medical illness, social isolation,
low-to-no income, and people of color. The Framework includes the specific poor outcomes that
people with pre-emergency social factors are more likely to experience at a disproportionately
higher level after emergencies: lack of access to post-incident services; displacement; injury,
illness, and death; property loss or damage; domestic violence; and loss of employment. I applied
the Social Determinants of Vulnerability Framework to the City of Boston to determine if the
relationships between social factors of vulnerability are consistent with literature, to determine
the areas of geographic concentrations of the most vulnerable people, and assess the relationship
between social isolation and the other factors of social vulnerability. Based on a citywide
geospatial analysis and neighborhood level correlation analyses, I also identified the most
vulnerable people and neighborhoods.
Key Words: Emergency Management, Social Vulnerability, Civil Rights, Social Isolation
5
TABLE OF CONTENTS
Introduction and Background ......................................................................................................... 9 Purpose .......................................................................................................................... 13
Methods......................................................................................................................... 14
Social Determinants of Vulnerability Framework ........................................................ 15
Social Vulnerability and Social Isolation ..................................................................... 16
Conclusion .................................................................................................................... 16
References ..................................................................................................................... 19
Paper 1: Social Determinants of Vulnerability Framework: Focusing Emergency Plans on the Needs of People in the American Cities ....................................................................................... 22
Introduction ................................................................................................................... 22
Problem Background .................................................................................................... 23
Purpose .......................................................................................................................... 24
Research Questions ................................................................................................... 26
Literature Review.......................................................................................................... 27
Social Vulnerability, Social Isolation, and the Impacts ............................................ 27
City Government Roles and Responsibilities ........................................................... 28
The Ignored Legal Imperatives ................................................................................. 30
Methodology ................................................................................................................. 31
Findings and Results ..................................................................................................... 33
Pre-Incident Attributes .............................................................................................. 34
6
Post-Incident Outcomes ............................................................................................ 36
Analysis and Synthesis ................................................................................................. 37
Emergency Management Core Capabilities .............................................................. 37
Legal Compliance and Assistance ............................................................................ 45
Next Steps and Future Research ............................................................................... 49
Conclusion .................................................................................................................... 51
Invest Now or Pay Later ........................................................................................... 52
References ..................................................................................................................... 53
Paper 2: Application of the Social Determinants of Vulnerability Framework to the City of Boston ........................................................................................................................................... 64
Background ................................................................................................................... 64
Social Determinants of Vulnerability Framework ........................................................ 66
Purpose .......................................................................................................................... 68
Methods......................................................................................................................... 68
Proxy Data ................................................................................................................ 69
Mapping .................................................................................................................... 70
Correlation and Regression Analysis ........................................................................ 71
Findings and Results ..................................................................................................... 71
Citywide Geographic Concentration of Social Determinants of Vulnerability Factors
............................................................................................................................................... 73
Social Isolation and Social Vulnerability ................................................................. 74
7
Social Determinants of Vulnerability within Boston Neighborhoods ...................... 75
Analysis and Recommendations ................................................................................... 83
Mitigation .................................................................................................................. 83
Response ................................................................................................................... 85
Recovery ................................................................................................................... 86
Legal Compliance and Assistance ............................................................................ 88
Future Research ............................................................................................................ 89
Conclusion .................................................................................................................... 90
References ..................................................................................................................... 93
Appendix A: Social Determinants of Vulnerability Frameworks ................................. 97
Appendix B: Neighborhood Correlation Analyses ..................................................... 112
Appendix C: Sum and Percentage of Neighborhood Populations by Social
Determinants of Vulnerability ................................................................................................ 130
References ................................................................................................................... 132
8
LIST OF TABLES AND FIGURES
Table 1. Social Determinants of Vulnerability Framework: Pre-Incident Social Factors ............ 67 Table 2. Social Determinants of Vulnerability Framework: Post-Incident Outcomes ................. 67 Table 3. Correlations for Socially Vulnerable Populations in the City of Boston ........................ 72 Table 4: Model of Social Isolation in Boston, MA ....................................................................... 75 Table 5. Neighborhood-Level Social Factors that Correlate with Social Isolation ...................... 77 Table B1: Sum of Neighborhood Populations ............................................................................ 130 Table B2: Percentages for Neighborhood Populations ............................................................... 131
9
Introduction and Background
For the first time in human history, more people across the world live in cities (Swiss
Reinsurance Company Ltd., 2013). Approximately 80.7% of the United States population lives
in metropolitan areas (U.S. Census Bureau, 2013). The growing concentration of people, assets,
and infrastructure in conjunction with the threats and hazards from natural, technological, and
human-caused events means that the loss potential in urban areas is high and continues to rise
(Swiss Reinsurance Company Ltd., 2013). This population density also means socially
vulnerable populations exist in higher numbers, further compounding risk in cities (Dwyer,
Zoppou, Nielsen, Day, & Roberts, 2004).
The social systems in cities are complex. People depend upon intricate social and
physical infrastructure, such as health and human services, public transportation, and utility
networks such as water, electricity and telecommunications (Dwyer et al., 2004). The potential
for poor outcomes after disasters in cities increases based on these complex systems, a higher
density of people, and larger numbers of socially vulnerable people (Galea, Freudenberg, &
Vlahov, 2005; Pelling, 2003). The daily circumstances of people are significant factors in cities’
ability to withstand the impact of an emergency (Intergovernmental Panel on Climate Change,
2012).
Social vulnerability is the susceptibility of social groups to the impacts of hazards such as
suffering disproportionate death, injury, loss, or disruption of livelihood; as well as their
resiliency, or ability to adequately recover from the impacts (Cutter & Emrich, 2006; Wisner &
Handmer, 1999). This susceptibility is a function of the demographic characteristics of the
population as well as more complex conditions such as health care provision, social capital, and
access to lifelines (Cutter & Emrich, 2006). Furthermore, at-risk populations have a higher
10
likelihood to be socially isolated, which has proven to be an indicator of increased mortality
before and after disasters (Klinenberg, 1999; Pantell et al., 2013). Socially vulnerable
populations are faced with a comparatively higher number of stressors before an emergency ever
happens (Gustafsson et al., 2014). However, if community and government services are equitable
and accessible before and after emergencies, socially vulnerable populations can have the same
opportunities as everyone else to be more resilient (Chandra et al., 2011). When socially
vulnerable populations are more resilient it increases the overall resilience of the city.
Researchers have identified many people as being socially vulnerable including those
who are children, older adults, people of color, low-income, living alone, single parents, non-
English speaking as well as those who suffer from chronic physical and mental illness,
disabilities, and low-literacy. Socially vulnerable populations have a disproportionate exposure
to risk and a decreased ability to avoid or absorb potential harm.
Emergency management planning identifies the actions local government will take
before, during, and after emergencies. The current process to develop plans focuses on reducing
the impact of emergencies on critical infrastructure, assets, and the environment. However, they
do not include ways to reduce the impact of emergencies on people. Therefore, efforts often
result in municipalities preparing for emergencies without accounting for the complex interaction
of social, physical, and hazard environments (U.S. Department of Homeland Security, 2010a).
Existing plans are designed for people who can walk, run, drive, see, hear, pay, and quickly
respond to directions (Greenberger, 2007; Kailes & Enders, 2007). The assumptions do not align
with the reality that at least half of the American population can be considered vulnerable to
disasters because of their existing social circumstances (Kailes & Enders, 2007). The approach
11
to emergency planning has to shift to incorporate the diverse needs of socially vulnerable people
into mitigation, response, and recovery.
Many people who are considered socially vulnerable are also protected by civil rights.
When there is a lack of inclusive planning, jurisdictions may be inadvertently violating civil
rights. Civil rights statutes and supporting federal guidance protect the rights of Americans so
that they are not denied the benefit from or participation in federally-funded programs and
activities on the basis of race, color, national origin, disability, age, economic status, or limited
English proficiency (Milligan & Company, 2007; Paulison, 2005; CDC, 2010). Cities are not
compliant with the protection afforded by these rights if these populations are not represented in
the planning process, included in the considerations for emergency management plans, or post-
incident services provided by local, state, federal, and nongovernmental organizations. The lack
of inclusion is de facto exclusion and results in local government developing and executing plans
that do not meet the needs of their constituents and potentially violates their civil rights.
There has been some action by emergency managers to address the needs of people with
disabilities, primarily in emergency sheltering and evacuation, as the result of Los Angeles and
New York City being sued for a lack of inclusive plans under the Americans with Disabilities
Act (Sherry & Harkins, 2011). I developed the Social Determinants of Vulnerability Framework
to identify the relationship between social factors that increase vulnerability to support inclusive
emergency planning. The social characteristics that are most interconnected include all of the
legally protected classes outlined above. Instead of waiting for lawsuits, cities can take action
now by using the Framework to identify the neighborhoods that have higher concentrations of
vulnerable populations and partner with them to work towards plans that meet their needs and
more efficiently manage limited resources. Public engagement in emergency management
12
provides people with a voice in decisions that impact them and fosters resilience (U.S.
Department of Health and Human Services, 2009). In fact, public engagement combats low-
decision latitude (lack of control over decisions made that impact them) which is one of the
major stressors linked to poor physical and mental health outcomes (Gustafsson et al., 2014).
Although people are not responsible for the occurrence of a natural disaster, we can
change the severity of the consequences (Abkowitz, 2008). The impact of a disaster on any
community is not random; it is determined by the daily circumstances of people living in the area
(Klinenberg, 2002). This means that local governments need to evaluate the interaction of
natural and manmade hazards, the living conditions of the city's most at-risk residents, and the
capabilities of local government (the organization most responsible for protecting the welfare of
residents) to determine the potential public health and safety consequences a disaster can inflict
(Klinenberg, 1999). Risk is an unavoidable reality in everything we do and it is not possible to
completely eliminate exposure to it. Instead, we have to understand the risk we are exposed to
and manage the exposure more effectively. The Social Determinants of Vulnerability Framework
can be integrated with existing risk assessments to better understand the characteristics of people
and neighborhoods that will likely suffer the most. It also provides a roadmap that can facilitate
community engagement to better understand needs of the community.
Academic literature and national guidance call for incorporating the considerations of
socially vulnerable populations into emergency planning. Understanding these social
vulnerabilities allows for inclusive plans and capabilities that can withstand and mitigate the
impacts of an incident and support community resilience (Chandra et al., 2011).
The complex interplay between social factors provides the most accurate picture of who
is more likely to experience higher exposure to post-incident impacts of emergencies (Levac,
13
Toal-Sullivan, & O'Sullivan, 2012). Further, understanding the intricacies of these social factors
prevents assumptions about the vulnerability of people because they fit into one category. This
approach also helps to focus limited resources on where they are needed most. However, the
literature provides a long and complicated list of social vulnerability characteristics and
conditions. What is needed is a practical, evidence-informed strategy for public health
preparedness and emergency management practitioners to understand how social factors are all
related. This will guide engagement of the right people in the planning process, develop inclusive
plans, and equitably implement those plans.
Purpose
This research focuses on developing a replicable, practical approach to understanding the
complexity of social vulnerability in American cities for policy makers and emergency
management practitioners across all sectors of government and industry, particularly public
health emergency preparedness. The study was conducted in two phases.
The first phase of the research identifies the co-existence of social vulnerability
categories and the social, physical, economic, and psychological health impacts of exposure to
hazards. The goal was to develop a Social Determinants of Vulnerability Framework that
focused on the characteristics of social vulnerability and their associated impacts. The
Framework helps to incorporate the diverse needs of socially vulnerable people into emergency
management planning through an inclusive planning process that supports self-determination,
respecting the fact that people are most knowledgeable about their needs. Emergency planners
can then apply this Framework to mitigation efforts such as risk assessments and community
resilience, as well as response and recovery efforts.
14
The Social Determinants of Vulnerability Framework can be applied to the unique
context of any city to explore the relationship between social factors. In this study, the Social
Determinants of Vulnerability Framework was applied to the City of Boston. The research
questions for both phases were: What are the socially vulnerable attributes that appear most
frequently in the literature? What are the relationships between these frequently occurring
attributes? Does the frequency and interrelationship represented in the literature exist in the City
of Boston? Which areas of Boston do we need to focus on for targeted mitigation, response, and
recovery planning?
Methods
In order to answer to the previous questions, this study uses a mixed methods approach in
two phases. First, a grounded theory approach was used to develop the Social Determinants of
Vulnerability Framework which shows the interrelationships between social factors to determine
which ones were most related to other social factors. Sixty-three social vulnerability attributes
and their relationships to other social factors were uploaded into TouchGraph Navigator, a link
analysis software. I used social network analysis logic to identify the relationships between each
social factor and associated social factors. This methodology identified co-occurrence and
frequency of co-occurrence across attributes. The result of this analysis is the Social
Determinants of Vulnerability Framework which depicts the co-existing socially factors on
which to focus mitigation, response, and recovery planning. The details of phase one are
represented in paper one which has been written for PLOS Currents: Disasters.
The second phase is the application of the Social Determinants of Vulnerability
Framework to the City of Boston. This phase is documented in paper two which was written for
the International Journal of Disaster Risk Reduction. For each social condition or characteristic,
15
the geospatial hot spots were identified. In order to compare the relationships in the literature to
the Boston data, a correlation analysis was conducted for social factors from the Social
Determinants of Vulnerability Framework at the city level and for each neighborhood. In order
to explore the relationship between social isolation and social vulnerability, I conducted a
regression analysis using social isolation as the dependent variable and the remaining social
factors from the Framework as the independent variables.
Social Determinants of Vulnerability Framework
The first phase of the research was a qualitative link analysis based on the literature.
There were seven pre-incident social factors that seemed to be driving social vulnerability based
on the number of links to other pre-incident social factors: children, people with disabilities,
older adults, chronic and acute medical illness, social isolation, low-to-no income, and people of
color. These seven social conditions and characteristics are directly or indirectly connected to
six post-incident outcomes that further increase social vulnerability after emergencies.
Post-incident conditions represent the types of exposure people experience after an
emergency. There were a total of eight post-incident outcomes from the literature. Six of the
eight had at least one link to pre-incident conditions. Lack of access to post-incident services and
displacement were related to the largest number of pre-incident social characteristics. These two
post-incident outcomes are directly related to three of the most frequently occurring pre-incident
conditions: social isolation, low-to-no income, and people of color. The other five post-incident
impacts are exposure to injury, illness, and death; property loss or damage; domestic violence;
and loss of employment.
Data collected for the seven pre-incident social conditions and characteristics in the
framework as well as five additional factors (limited English proficiency, renters, less than high
16
school, women, and lack of vehicle) were analyzed for the City of Boston. In the City of Boston,
the neighborhoods that are likely to have the highest risk of poor outcomes were Mattapan,
Roxbury, and South Dorchester. These three neighborhoods had multiple census tracts with hot
spots for social isolation, low-to-no income, and people of color which were all associated with
the post-incident outcomes in the Social Determinants of Vulnerability Framework. Additionally,
in East Boston, Hyde Park, Mattapan, and Roxbury, social isolation is correlated with all of the
social determinants of vulnerability. This indicates that poor outcomes after emergencies are
likely. These neighborhoods deserve attention to their unique conditions by emergency
management practitioners.
Social Vulnerability and Social Isolation
During both phases of this research, social isolation was found to be a consistent
underlying social factor among the factors of social vulnerability. Social isolation has been
validated as being driven by social vulnerability via link analysis based on the literature,
correlation analysis (citywide and within neighborhoods), and regression analysis.
The regression analysis based on the census tract data for the City of Boston confirmed
the significance of the relationship between the social determinants of vulnerability and social
isolation. People with disabilities, children, older adults, low-to-no income, less than high school
education, people of color, women, and renters explained over 95 percent of the variation in
social isolation. People with limited English proficiency were highly correlated with each of the
eight social factors listed above.
Conclusion
The results of this research warrant a number of policy changes. Cities have to rethink
emergency management approaches to mitigation, response, and recovery planning and
17
implementation. The Social Determinants of Vulnerability Framework provides a tool that cities
can use to identify areas with higher concentrations of social factors that increase vulnerability,
develop relationships and inclusive emergency plans, and equitably execute those plans. A key
benefit of the Framework is that it includes all of the legally protected classes of people. By
including them in emergency planning, cities can reduce the likelihood of civil rights violations.
The Federal Emergency Management Agency has provided some guidance for those
receiving federal funding on protecting the civil rights after emergencies to prevent people being
denied or excluded from federally funded programs on the basis of race, color, national origin,
disability, age, economic status, or limited English proficiency. The U.S. Department of Health
and Human Services (HHS), which includes the Center for Disease Control and Prevention
(CDC), has also provided guidance on laws that are important during emergencies that affect
public health. However, the guidance does not include an explanation of the civil rights laws or
executive orders that protect people’s right to equitable access before, during, and after disasters.
Response efforts such as public information and warning, sheltering, and evacuation, are
emergency management programs offered by federally funded activities that must also prevent
exclusion for these protected classes. This includes mitigation efforts such as risk assessment,
community preparedness and resilience, and long term vulnerability reduction. Mitigation is
especially important because the actions taken before an emergency reduce the level of response
and recovery needed after an emergency.
The Social Determinants of Vulnerability Framework provides FEMA, HHS, and
other federal agencies with an approach to inform guidance to cities that can improve their
ability to meet the diverse needs of people during all phases of an emergency. The Framework
offers cities a proactive approach to determine who is most vulnerable and take action to develop
18
inclusive emergency plans. The following two publishable papers provide further details on the
Social Determinants of Vulnerability Framework, its application to the City of Boston, and
recommendations for policy and strategy changes.
19
References
Abkowitz, M. D. (2008). Opertional Risk Management: A Case Study Approach to Effective
Planning and Response. Hoboken, NJ: John Wiley & Sons, Inc.
Chandra, A., Acosta, J., Stern, S., Uscher-Pines, L., Williams, M. V., Yeung, D., . . . Meredith,
L. S. (2011). Building Community Relience to Disasters: A Way Forward to Enhance
National Health Security. Santa Monica: RAND Corporation.
Cutter, S. L., & Emrich, C. T. (2006). Moral Hazard, Social Catastrophe: The Changing Face of
Vulnerability along the Hurricane Coasts. Annals of the American Academy of Political
and Social Science, 604(1), 102-112.
Dwyer, A., Zoppou, C., Nielsen, O., Day, S., & Roberts, S. (2004). Quantifying Social
Vulnerability: A methodology for identifying those at risk to natural hazards. Geoscience
Australia Record, 2004/14.
Galea, S., Freudenberg, N., & Vlahov, D. (2005). Cities and population health. Social Science &
Medicine, 60(5), 1017-1033. doi: http://dx.doi.org/10.1016/j.socscimed.2004.06.036
Greenberger, M. (2007). Preparing Vulnerable Populations for Disaster: Inner-city Emergency
Preparedness - Who Should Take the Lead. Journal of Health Care Law, 291-308.
Gustafsson, P. E., San Sebastian, M., Janlert, U., Theorell, T., Westerlund, H., & Hammarström,
A. (2014). Life-Course Accumulation of Neighborhood Disadvantage and Allostatic
Load: Empirical Integration of Three Social Determinants of Health Frameworks.
American Journal of Public Health, 104(5), 904-910.
Intergovernmental Panel on Climate Change. (2012). Managing the Risks of Extreme Events and
Disasters to Advance Climate Change Adaptation. In C. Field, V. Barros, T. Stocker, Q.
Dahe, D. J. Dodden, K. L. Ebi, M. D. Mastrandrea, K. Mach, G.-K. Plattner, S. K. Allen,
20
M. Tignor & P. Midgley (Eds.), Special Report of the Intergovernmental Panel on
Climate Change. New York.
Kailes, J. I., & Enders, A. (2007). Moving Beyond "Special Needs". Journal of Disability Policy
Studies, 17(4), 230-237.
Klinenberg, E. (1999). Denaturalizing disaster: A social autopsy of the 1995 Chicago Heat
Wave. Berkeley: University of California, Berkeley.
Klinenberg, E. (2002). Heat wave: a social autopsy of disaster in Chicago. Chicago: The
University of Chicago Press.
Levac, J., Toal-Sullivan, D., & O'Sullivan, T. L. (2012). Household emergency preparedness: a
literature review. J Community Health, 37(3), 725-733. doi: 10.1007/s10900-011-9488-x
Milligan & Company, L., ,. (2007). Transportation Equity in Emergencies: A Review of the
Practices of State Departments of Transportation, Metropolitan Planning Organizations,
and Transit Agencies in 20 Metropolitan Areas (Vol. 2014).
Pantell, M., Rehkopf, D., Jutte, D., Syme, S. L., Balmes, J., & Alder, N. (2013). Social Isolation:
A Predictor of Mortality Comparable to Traditional Clinical Risk Factors. American
Journal of Public Health, 103(11), 2056-2062.
Paulison, R. D. (2005). Civil Rights Program. Washington, D.C.: U.S. Department of Homeland
Security.
Pelling, M. (2003). Vulnerability in Cities. London: Earthscan Publications.
Sherry, J. M., Nishamarie, & Harkins, J. D., Anne Marie. (2011). Leveling the emergency
preparedness playing field. Journal of Emergency Management, 9(6), 11-16. doi:
10.5055/jem.2011.0075
21
Swiss Reinsurance Company Ltd. (2013). Mind the Risk: A global ranking of cities under threat
from natural disasters. Zurich.
U.S. Census Bureau (Producer). (2013, August 11). Growth in Urban Population Outpaces Rest
of Nation, Census Bureau Reports. Census.gov. Retrieved from
http://www.census.gov/newsroom/releases/archives/2010_census/cb12-50.html
U.S. Centers for Disease Control and Prevention. (2010). Public Health Workbook: To Define,
Locate, and Reach Special, Vulnerable, and At-Risk Populations in an Emergency.
Washington, D.C.
U.S. Department of Health and Human Services. (2009). National Health Security Strategy of
the United States of America. Washington, D.C.
U.S. Department of Homeland Security. (2010a). Comprehensive Preparedness Guide 101:
Developing and Maintaining Emergency Operations Plans. Washington, D.C.
Wisner, B., & Handmer, J. (1999). Hazards, globalisation, and sustainability. Paper presented at
the Development in Practice.
22
Paper 1: Social Determinants of Vulnerability Framework: Focusing Emergency Plans on the Needs of People in the American Cities
“Disasters fracture us along fault lines that already exist.” Jim Siemianowski, LICSW
Introduction
Over 80 percent of the U.S. population lives in cities that are dependent upon intricate
social and physical infrastructure, such as health and human services, public transportation, and
utility networks, such as water, electricity and telecommunications (Dwyer, Zoppou, Nielsen,
Day, & Roberts, 2004; Molin Valdes, Rego, Scott, & Aguayo, 2012; U.S. Census Bureau, 2013).
The potential for poor outcomes after disasters in cities is incredibly high as the result of the
complex infrastructure, higher density of people, and large numbers of socially vulnerable
populations (Galea, Freudenberg, & Vlahov, 2005; Pelling, 2003).
Experience has shown that individuals who can help themselves during and after a
disaster will usually do so. However, a disaster strains the limited capability and capacity of
some populations to effectively respond and recover. Their circumstances reduce their ability to
prepare for, cope with, and adapt to the impact of emergencies. Socially vulnerable people have
existing social circumstances generally associated with age, gender, race, family composition,
medical illness, disability, literacy, English proficiency, and social isolation (CDC, 2010).
Since 9/11, local municipalities have invested considerable time and financial resources
into planning for disasters. In spite of this, local governments have made little progress in
developing inclusive emergency plans to address the needs of all residents. In order to be fully
prepared, local governments need to develop inclusive emergency plans that address the needs of
people who will likely need more assistance.
23
Problem Background
There are many barriers to inclusive preparedness: (1) There is the false notion that
preparedness is exclusively an individual responsibility, (2) cities do not have clear guidance on
how to achieve inclusive planning, (3) there is limited understanding of the way social
vulnerability factors are related to one another, and (4) there is a limited understanding of civil
rights protections afforded to certain socially vulnerable populations.
There is a perception that preparedness and resilience is solely a personal trait (Mohaupt,
2009). This attitude leads to lecturing people on all the plans, supplies, knowledge they need to
have regardless of their capacity or capability to obtain them. It leads to passing judgment only
on people and not holding the local government accountable for their role. For example,
practitioners often question why people in New Orleans did not evacuate before Hurricane
Katrina. They knew it was coming, right? However, research has shown that many people did
not evacuate because the messages warning them were not clear and accessible; or they did not
have the resources or physical ability to leave (Milligan & Company, 2007). Socially vulnerable
populations are often not in a position to do all of the things emergency management expects of
them without assistance. Further, the assistance local jurisdiction provides has to reach them and
meet their actual needs.
Federal guidance and research indicates that resilience is a shared process among
individuals, communities, and government. The national planning frameworks for mitigation,
response, and recovery identify local government as being ultimately responsible for building
preparedness in partnership with the community. City leadership has to deal with the impact of
emergencies regardless of how large or small (Molin Valdes et al., 2012). There is an
opportunity for cities to partner with constituents, particularly vulnerable populations, to develop
24
informed plans that will meet their needs and improve physical, mental, and economic health
outcomes.
Another reason local emergency planners are limited in their ability to develop inclusive
plans is because they are overwhelmed with long lists of people who may be considered
vulnerable. If so many people are considered vulnerable, where and how does a practitioner
focus on those who are most vulnerable? At the other end of the spectrum, some practitioners
have taken a hyper-focused approach and have limited their inclusive planning efforts to a
narrow group of people. In many cases, the group they tend to focus on is people with
disabilities. This choice is often made because other cities have been sued for not including
people with disabilities. Also, there has been a significant amount of guidance from disability
advocates and the federal government on how to accommodate people with disabilities. The
result of exclusive planning occurs whether there is a laundry list of social vulnerability factors
that cannot be operationalized or cities are focusing on a very specific group of people.
Each category of social vulnerability presents challenges for people. However, it is the
interaction of these social factors that intensifies vulnerability. This overlap exponentially
increases the level of exposure to risk and suffering such as injury, death, illness, and difficult
recoveries (Morrow, 1999). Many of the categories of people considered socially vulnerable are
protected by multiple civil rights laws and federal guidance. However, local jurisdictions do not
seem to understand that these laws apply to all of their emergency management activities, from
mitigation to recovery.
Purpose
The goal of this project is to identify individuals with the greatest vulnerability to
disasters and to help emergency management and public health preparedness practitioners to take
25
appropriate actions during the mitigation, response and recovery phases of emergencies. In order
to better understand the social characteristics that increase vulnerability, I developed the Social
Determinants of Vulnerability Framework. The Framework is based on a link analysis of social
factors of vulnerability found in the literature.
The link analysis identified seven pre-disaster social factors: children, people with
disabilities, older adults, chronic and acute medical illness, social isolation, low-to-no income,
and people of color. While each factor alone may present challenges for individuals, the
interaction of these social factors intensifies vulnerability. People with these pre-disaster factors
are more likely to be exposed to post-incident outcomes such as injury, illness, and death;
displacement; limited access to post-emergency services; domestic violence; loss of
employment, and property damage.
Local emergency planners can use the Framework to: (1) map pre-incident social factors
and determine the high concentration areas and include it into risk assessments; (2) understand
how each of the factors are related to others; (3) develop outreach plans to partner with those
communities, (4) conduct an inclusive emergency planning process; and (5) equitably execute
the plans in the response and recovery phases based on the links between social vulnerability and
post-incident outcomes. By using the Social Determinants of Vulnerability Framework, local
emergency management and public health preparedness planners can reduce the potential public
health and safety consequences of these disasters.
Since the Framework includes categories of people explicitly protected by civil rights
laws, local governments can reduce the likelihood of infringing on the rights of those groups
during mitigation, response, and recovery. Further, local government will uphold the fourth pillar
26
of public administration, social equity in services, which means the people who have the greatest
need receive the requisite services (Norman-Major, 2011).
Without the proper analysis, it is easy to make false assumptions about the characteristics
of people because they belong to one of the at-risk groups. Unfortunately, this is the current state
of the discussion on social vulnerability in emergency management. Members of individual
population groups are not equally vulnerable, nor are they merely victims. They are part of
communities with many strengths that can support inclusive planning. There are many
circumstances that enable people to assist in some situations but require assistance in others
(Flanagan, Gregory, Hallisey, Heitgerd, & Lewis, 2011). Social vulnerability is not static and can
be reduced if we commit to engage and strengthen institutional and individual capacity to cope
and act to reduce risk (Molin Valdes et al., 2012).
This research does not focus on specific types of hazards or identifying the areas that are
exposed to hazards. There has been a significant amount of research on hazards and their
impacts. Social vulnerability exists in spite of hazards or where they impact. Planners can use the
Social Determinants of Vulnerability Framework to inform existing risk and hazard analyses and
include the needs of people exposed to hazards into the planning process.
Research Questions
This research focuses on identifying the co-existence of social vulnerability categories
and the social, physical, economic, and psychological health impacts of emergencies. The
research questions are: What are the relationships between the characteristics of social
vulnerability? Which ones have the highest number of links to other characteristics of social
vulnerability? The result of the analysis was used to develop the Social Determinants of
Vulnerability Framework to inform mitigation, response, and recovery efforts to ensure the
27
diverse needs of socially vulnerable people are incorporated into emergency management
planning.
Literature Review
Existing literature does not take into account the manner in which social vulnerability
factors are often compounded to produce negative consequences associated with high risk. The
sheer volume and unclear relationships among social factors becomes a practical challenge to
identifying vulnerable populations within a community and developing strategies to reduce their
exposure to public health and safety consequences of emergencies.
Social Vulnerability, Social Isolation, and the Impacts
Social vulnerability is the result of pre-emergency social factors that create a lack of
capacity or capability to prepare for, response to, and recover from emergencies. Social
vulnerability includes people who are more likely to suffer disproportionately because of their
existing social circumstances such as those associated with age, gender, race, medical illness,
disability, literacy, and English proficiency, and social isolation (CDC, 2010). Their
circumstances increase the likelihood of social isolation, which is a lack of engagement in social
ties, institutional connections, or community participation (Pantell, Rehkopf, Jutte, Syme,
Balmes, & Adler, 2013; Seeman, 1996). Social isolation in daily life or post-disaster is directly
correlated with higher mortality (Klinenberg, 2002; Pantell, Rehkopf, Jutte, Syme, Balmes, &
Alder, 2013).
Socially vulnerable people experience high levels of adversity in their daily lives. The
higher amount of stressors on a regular basis significantly increases the use of physiological
responses that wear the body down over time (Logan & Barksdale, 2008). The sum of exposure
to chronic stressors over time and repeated heightened physiological response is referred to as
28
allostatic load (Ganzel, Morris, & Wethington, 2010). Allostatic load is a framework for the
collective cognitive and physical deterioration the body experiences because of continual
exposures to stressors during the life course. The more stress people experience the quicker and
more significantly their physical and mental health is worn down. There is increasing evidence
that the cost to the body from allostatic load interferes with people’s ability to adapt to future
stressors (Ganzel et al., 2010). Additionally, allostatic load has been considered the biological
link that explains disparities in mortality and morbidity based on social conditions and
characteristics (Gustafsson et al., 2014).
Chronic stressors cumulatively reduce the physical and psychosocial resilience of
vulnerable people in our communities. An acute stressor, such as a disaster, can deplete any
remaining physical and psychosocial resilience. Additionally, when interactions with the
institutions that are supposed to be there to help are not supportive, there is an immediate
physical reaction that contributes to poor health (Seeman, 1996). Without consideration for
social conditions of communities in emergency plans, municipalities contribute to the problem
and cause unnecessary additional suffering and poor recovery outcomes. However, there are
protective factors that help people cope with adversity and reduce social isolation. Protective
factors include building neighborhood social connections and improving access to government,
community, and private services before and after an emergency (Mohaupt, 2009).
City Government Roles and Responsibilities
The National Response Framework indicates that “[m]ost incidents begin and end at the
local level” (U.S. Department of Homeland Security, 2013b, p. 6). In fact, the National
Mitigation Planning Framework has identified local governments as having the largest number of
roles and responsibilities that advance mitigation (U.S. Department of Homeland Security,
29
2013a). The National Disaster Recovery Framework states that leadership is provided locally for
any support from federal agencies during recovery efforts (Federal Emergency Management
Agency, 2011). In all emergency management mission areas, local government is ultimately
responsible for emergency planning to meet the diverse needs of people. However, local
governments have been slow to engage diverse communities and incorporate their needs into
mitigation, response, and recovery planning.
Inclusive planning is also a matter of national health security (U.S. Department of Health
and Human Services, 2009). The National Health Security Strategy focuses on the nation’s goal
to protect people’s health in the case of any incident that puts health and well-being at risk.
When enacted in December 2006, Pandemic and All Hands Preparedness Act required the U.S.
Department of Health and Human Services (HHS) “to integrate the needs of at-risk individuals
on all levels of emergency planning, ensuring the effective incorporation of at-risk populations
into existing and future policy, planning, and programmatic documents.” National emergency
management and public health preparedness guidance emphasize leadership at the local
government level and the importance of inclusive planning.
However, existing emergency management efforts often result in municipalities preparing
to support a homogenous community during large-scale emergency or disaster without
accounting for the complex interaction of social, physical, and hazard environments (U.S.
Department of Homeland Security, 2010a). At a minimum, emergency plans acknowledge some
of these populations exist. However, there is a lack of clear explanation on how emergency
management plans will address the needs of vulnerable populations. These plans are designed for
people who can walk, run, drive, see, hear, pay, and quickly respond to directions (Greenberger,
2007; Kailes & Enders, 2007). The assumptions do not align with the reality that at least half of
30
the American population could be considered vulnerable to disasters because of their existing
social circumstances (Kailes & Enders, 2007).
The Ignored Legal Imperatives
Although this research project does not focus on the legal aspects of social vulnerability
and emergency management, it is an important component of policy and planning realities.
Emergency managers in cities have been motivated to take action to make reasonable attempts to
accommodate people with disabilities out of fear that they will be sued based on the Americans
with Disabilities Act violations. This fear is based on Los Angeles and New York City being
sued for lack of ADA compliance in their emergency plans (Sherry & Harkins, 2011). However,
there are legal imperatives for other social characteristics and conditions. Cities should not wait
for civil rights advocates to bring suit in order to begin to include the diverse needs of people in
emergency plans. Precedent or prevalence of lawsuits should not ggyuide adherence to
accommodating the needs of the communities we serve. A healthy respect for the letter and spirit
of the law should be enough.
Civil rights statutes, including Title VI of the Civil Rights Act of 1964, Section 504 and
508 of the Rehabilitation Act of 1973, the Age Discrimination Act of 1975, the 1968 Fair
Housing Act, and Sections 308-309 of the Robert T. Stafford Disaster Relief and Emergency
Assistance Act of 1988 (as amended) provide that persons in the United States shall not be
denied the benefits of, excluded from participation in, or subject to discrimination under
federally funded programs or activities on the basis of race, color, national origin, disability, age,
or economic status (Milligan & Company, 2007; Paulison, 2005). Additionally, following the
August 11, 2000 passage of Executive Order 13166, “Improving Access to Services for Persons
with Limited English Proficiency,” people with limited English proficiency qualify for the same
31
anti-discrimination protection designated for race, color, or national origin under Title VI of the
Civil Rights Act (U.S. Centers for Disease Control and Prevention, 2010).
Protecting the rights of all people extends through the recovery phase. Section 308 of the
Robert T. Stafford Emergency Management and Disaster Assistance Act reinforces the
prohibition of discrimination on the basis of race, color, religion, disability, nationality, sex,
English Proficiency, age, or economic status with regard to disaster assistance programs. Rights
relevant to recovery are also protected under the 1968 Fair Housing Act. Despite the existence of
these laws, there have been documented violations of civil rights statutes after disasters (Muñiz,
2006). The provision of accessible and appropriate recovery services is not charity, but a human
right recognized by the United Nations (Brookings-Bern Project on Internal Displacement, 2008)
and protected by the previously mentioned laws and directives. Social equity is one of the four
pillars of public administration but is still not practiced as a standard for the manner in which the
government provides all services (Norman-Major, 2011), including mitigation, response, and
recovery in emergency management.
The goal of developing the Social Determinants of Vulnerability Framework is to identify
the interrelationships of social factors that increase vulnerability to better understand the needs of
the community, focus limited resources on those who need them most, and help protect the rights
of all people cities are supposed to serve.
Methodology
The data used were based on existing social vulnerability literature, which still has some
gaps in the populations that have been studied. For example, there was limited literature
regarding the lesbian, gay, bisexual, and transgendered populations and their increased
vulnerability as it relates to other social conditions.
32
The Framework was developed using link analysis or network analysis to facilitate a
grounded theory approach. This process is similar to social network analysis and is beneficial
because of its ability to reveal patterns in complex data that would be undetectable using other
analytic approaches (Knobel, 2013). The co-occurrence of social factors across the literature was
modeled using the TouchGraph Navigator, a link analysis software, to identify the frequency and
strength of relationships between variables.
The literature indicated that vulnerabilities exist based upon pre-incident social
circumstances. Some of the literature also provided insight into the post-incident outcomes from
disaster exposure that socially vulnerable people are more likely to face because of their existing
social circumstances. The initial categories used to identify pre-incident variables included age,
race, income, household composition, family composition, housing type, disease/illness, access,
language and literacy, non-residents, gender, and disability (Cutter & Emrich, 2006; Kailes &
Enders, 2007; U.S. Centers for Disease Control and Prevention, 2010). These categories helped
to guide the literature review to compile a list of 63 social vulnerability attributes relevant to
cities.
Categories for post-incident outcomes included exposure to injury, death, illness,
property damage, losing a love ones, losing a business, or limited access to recovery services
(Isaranuwatchai, Coyte, McKenzie, & Noh, 2013). These outcomes were extracted from a
literature review and provide an organizational structure that can capture the multiple categories
of vulnerability in to which a person can fall (See Appendix A for a full list).
Each of the 63 social attributes or factors was researched to identify in the literature the
related social characteristics that (1) increase vulnerability or (2) often coincide with the 63
social factors. Understanding the co-occurrence of socially vulnerable characteristics was an
33
inductive process. The measures and respective attributes evolved based on the results of the
literature review. The data collected for each attribute was organized in an Excel spreadsheet
utilizing a codebook and inductive reasoning (see Appendix B). The spreadsheet was uploaded
into the TouchGraph Navigator software to identify the relationships between all of the
categories from the collected data.
The link analysis resulted in four types of nodes that represented the 63 factors of social
vulnerability with which I started, the sources for each of the social conditions, the subsequent
characteristics related to those attributes, and sources for the related characteristics (See
Appendix C). The relationships or links between the previously mentioned nodes were primary
attribute to primary attributes sources (the relationship between the 63 factors and the sources of
the data that identified them as socially vulnerable), associated characteristics to associated
attributes sources (the relationship between the social factors associated with the 63 primary
attributes and their sources), and primary attributes to associated characteristics (the relationship
between the 63 social factors and associated social factors).
This methodology identified co-occurrence and frequency of co-occurrence across
attributes. The result of the literature review and link analysis is the Social Determinants of
Vulnerability Framework which depicts (1) the co-existing, pre-incident socially vulnerable
characteristics and (2) the associated post-incident outcomes. (See Appendix D)
Findings and Results
There were seven pre-incident social factors that seemed to be driving social vulnerability
based on the number of links to other pre-incident factors. These seven social conditions are
directly or indirectly connected to the six post-incident outcomes that perpetuate social
vulnerability. A key concept to keep in mind is that people in any one category are not
34
necessarily vulnerable. Based on this analysis, it is primarily the presence of social isolation in
conjunction with any of the other categories that increases vulnerability.
Pre-Incident Attributes
Pre-incident social conditions represent the existing social vulnerability of people in
cities. These social factors are based on a review of purposively selected literature regarding
social vulnerability primarily in the context of emergencies or health. There were a total of 63
social factors. The social characteristics that had ten or more associated social factors became
part of the framework: chronic and acute medical illness, people of color, low-to-no income,
children, older adults, people with disabilities, and social isolation. These seven social factors,
outlined below, appear in the resultant Social Determinants of Vulnerability Framework.
Chronic and Acute Medical Illness. The most socially vulnerable people with chronic or
acute medical illnesses were low-to-no income older adults with a disability. Based on the
literature, the following categories also increased the vulnerability of people with chronic and
acute medical illnesses: alcohol dependency, assisted living facilities, drug dependency, group
homes, homebound, homeless, nursing home, psychological illness, residential care facilities,
and social isolation.
People of Color. People of color were linked to ten other social factors. People of color
who are most vulnerable are socially isolated and low-to-no income. Furthermore, people of
color with those characteristics are more likely to experience displacement after an emergency.
The literature also identified vulnerable people of color being associated with: children, lack of
health insurance, and multi-story buildings or multi-unit buildings. People of color was a
category that the literature also referred to as minorities and often times was inconsistent with
specifying specific categories or generalizing across all people of color, so I combined all
35
instances into people of color while maintaining the individual categories to capture any nuances
in considerations for specific communities.
Low-to-No Income. People with no-to-low income were also linked to all other social
conditions in the Social Determinants of Vulnerability Framework. This indicates that low-to-no-
income is a characteristic that consistently compounds risk. Furthermore, it was also linked to
homelessness; lack of healthcare; lack of vehicle; lesbian, gay, bisexual, and transgender people;
being a primary/sole caregiver or single parent; and women. The low-to-no income data
represents a calculated field that combines those who are 100 percent below the poverty level
and those who are 100 to 149 percent of the poverty level.
Children. Although children had many social characteristics associated with it (12), it
had the least that were linked to other social factors in the Framework. The children who seemed
to be most vulnerable were socially isolated, low-income, limited English proficient, and were
people of color. It should be noted that children are particularly vulnerable because in addition to
their own circumstances, they are also impacted by the circumstances of the adults providing for
their care (Shi & Stevens, 2010).
Older Adults. Older adults were most vulnerable when they were socially isolated, low-
to-no income, and had a disability. There were a total of 16 factors of social vulnerability
associated with older adults. Most of the literature considered older adults 65 and older.
People with Disabilities. The most vulnerable people with disabilities were those who
were older adults. Like people of color, people with disabilities was a category that the literature
often times was inconsistent with specifying categories or generalizing across all types of
disabilities, so I combined all instances into people with disabilities and maintained the
individual categories to capture any nuances in considerations for specific types of disabilities.
36
Social Isolation. Social isolation refers to a lack of engagement in social ties,
institutional connections, or community participation (Pantell, Rehkopf, Jutte, Syme, Balmes, &
Adler, 2013; Seeman, 1996). Social isolation had the largest number of links to other factors of
vulnerability. It was connected to all other social determinants of vulnerability in the Framework
through multiple social factors. This indicates that it has the highest centrality to all socially
vulnerable populations. Based on the literature and the results of this analysis, social isolation is
the most consistent contributor to social vulnerability.
Post-Incident Outcomes
Socially vulnerable people are disproportionately exposed to the daily adversities. This
constant exposure to stressors deteriorates their physical and cognitive health resulting in a
comparatively higher allostatic load. Allostatic load is the sum of the body’s reactions to
stressful events. The types of exposures people experience in their lives before an emergency
that increase allostatic load include: illness directly or to a parent or caregiver; residential
instability (this includes displacement); death or illness of a close loved one; social isolation;
limited opportunity to make their own decisions (low-decision latitude); and threat or violence
(Gustafsson et al., 2014). The results of this research indicate that many of the same stressors
people are exposed to before an emergency that increase poor physical and cognitive health
outcomes are the same as the stressors socially vulnerable people are likely to face after an
emergency.
Post-incident outcomes represent the types of impacts from an emergency or disaster
people may experience. There were a total of eight post-incident outcomes from the literature.
Six of the eight had at least one link to pre-incident social conditions: access to post-incident
services; displacement; injury, illness, and death; loss of employment; property damage; and
37
domestic violence. These post-incident consequences were directly or indirectly related to all of
the pre-incident social factors in the Framework. However, they were most significantly related
to three of them: social isolation, low-to-no income (had the most links to post-incident
outcomes), and people of color.
Analysis and Synthesis
The Social Determinants of Vulnerability Framework provides cities with a way to base
mitigation, response, and recovery mission areas on the needs of people in their jurisdiction.
Local emergency management analysis, planning, decision-making, and assignment of available
resources must be equitable and respect the human rights of constituents. Exclusion of socially
vulnerable populations in any of the emergency management mission areas, whether intentional
or not, may be a violation of civil rights. The Social Determinants of Vulnerability Framework
supports urban areas to reduce the likelihood of excluding the most vulnerable populations while
respecting the complexity of their social circumstances.
Emergency Management Core Capabilities
The Federal Emergency Management Agency has identified core capabilities that apply
to prevention, protection, mitigation, response, and recovery. Federal, state, and local entities are
supposed to be able to collectively build and deliver these capabilities in partnership with
communities (U.S. Department of Homeland Security, 2011). The focus of this research has been
on the capabilities associated with mitigation, response, and recovery.
Mitigation. The U.S. Department of Homeland Security has included seven core
capabilities in the mitigation mission area. I am going to focus on the ones most related to the
Social Determinants of Vulnerability Framework: risk and disaster resilience assessment,
community resilience, and long term vulnerability reduction. Mitigation is “the thread that
38
permeates the fabric of preparedness” and is intended to minimize the risks associated with
threats and hazards (U.S. Department of Homeland Security, 2013a, p. 6). Effective mitigation
starts with a risk assessment to identify the threats and hazards a community faces and determine
the associated vulnerabilities and consequences (U.S. Department of Homeland Security, 2013a).
Assessing risk and disaster resilience allows decision makers, responders, and community
members to take informed action to reduce their risk and increase their resilience (U.S.
Department of Homeland Security, 2013a).
Vulnerability is a factor of risk representing the susceptibility of a community to the
impact of hazards as determined by four domains: social, physical, economic, and environmental
(Chandra et al., 2011; United Nations Office for Disaster Reduction, 2005). Social vulnerability
is the susceptibility of social groups to the impacts of hazards such as suffering disproportionate
death, injury, loss, or disruption of livelihood; as well as their resiliency, or ability to adequately
recover from the impacts (Cutter & Emrich, 2006; Wisner & Handmer, 1999). This
susceptibility is a function of the demographic characteristics of the population as well as more
complex conditions such as health care provision, social capital, and access to lifelines (Cutter &
Emrich, 2006). However, social vulnerability is often not a consideration as part of traditional
risk assessments, which inform all emergency management mission areas. Consequently, the
most foundational aspect of mitigation, response, and recovery begins with excluding
considerations for how social vulnerability will be addressed. So a key question is how do we
incorporate social vulnerability into risk, which is the foundation of mitigation?
The United Nations (UN) has promoted a promising risk-based framework for making
cities more resilient (Dickson, Baker, & Hoornweg, 2012) that is consistent with other risk
frameworks. This approach to risk is promising because it includes an often forgotten element, a
39
socioeconomic assessment. The incorporation of this type of assessment includes the social
dimensions of vulnerability. The social conditions in the Social Determinants of Vulnerability
Framework can be used as the factors considered for assessing the social dimension of risk in
American cities. The Framework identifies the relationship between the pre-incident social
conditions and the post-incident outcomes. We cannot control all the reasons people are socially
vulnerable, the occurrence of all disasters, or all of the post-incident suffering. However, we can
affect social isolation, which is the most central socially vulnerable factor.
Social isolation is the product of a lack of social justice and social capital which are both
important aspects of resilience (Chandra et al., 2011). Social justice, social equity, and social
capital are related concepts. Social justice means the institutions serving the community enable
them to contribute to decisions about their community and prevent inequality. Social equity
means people get what is right for them. Social capital is the relationships people have with each
other, their community, and institutions.
Practical approaches to improving community preparedness, and therefore resilience,
include connecting socially vulnerable people with others in their communities as well as
government and community organizations that provide services meant to improve their well-
being and quality of life. This is one of the reasons involving public health preparedness is key to
successful emergency preparedness efforts: Every day, urban public health focuses on
connecting vulnerable populations with resources to promote their overall health and well-being.
All of the previously-mentioned actions represent increased efforts to move towards social equity
in urban emergency management and are key elements of long-term vulnerability reduction.
Thus, social justice and social capital are fostered to reduce the negative public health impact of
social vulnerability and overall risk before, during, and after emergencies (Durant, 2011).
40
Community resilience is an end state of effective risk management (U.S. Department of
Homeland Security, 2013a), which is a foundational component of mitigation. The key aspects of
community resilience are leadership, collaboration, partnership building, and education and skills
building (U.S. Department of Homeland Security, 2013a). Cities can lead collaborations with
community partners to customize community education and training based on the populations in
different neighborhoods. Instead of lecturing about what people should have and do, cities and
community partners can provide opportunities for diverse populations to work together to
collectively have supplies and develop plans within their capability and capacity. By focusing on
the resilience of the whole community, especially those who are most socially vulnerable, the
community’s adaptive capacity to recover from all types of change is enhanced regardless of the
threat or hazard. The Social Determinants of Vulnerability Framework can support local planners
in focusing risk and resilience efforts on being inclusive and accounting for the complexity of the
social conditions of people thereby contributing to long-term vulnerability reduction.
Response. Key core capabilities within the response mission area that have the most
implications for socially vulnerable populations are public information and warning, mass care
services, and critical transportation. “[T]raditional methods of communicating health and
emergency information often fall short of the goal of reaching everyone in a community” (U.S.
Centers for Disease Control and Prevention, 2010, p. 4). Public information and warning is one
of the national core capabilities and has three distinct elements: quality (coordinated, prompt,
reliable, and actionable, clear, consistent), accessibility (accessible, culturally and linguistically
appropriate methods), and purpose or content (information regarding any threat or hazard,
actions being taken, and the assistance being made available, as appropriate) (U.S. Department
of Homeland Security, 2011). The Social Determinants of Vulnerability will help local
41
government customize messaging to target people in their community normally excluded from
important communications. Targeted outreach and messaging has been successfully employed by
the private sector and political campaigns to customize messages and modes of communication
based on the target groups they are trying to reach (Issenberg, 2012).
Mass care services is the capability to provide “…life-sustaining services to the affected
population with a focus on hydration, feeding, and sheltering to those who have the most need,
as well as support for reunifying families” (U.S. Department of Homeland Security, 2011, p. 13).
Additionally, Functional Needs Support Services (FNSS) enable individuals to maintain their
independence in a general population shelter (Federal Emergency Management Agency, 2010).
Mass cares services should be prepared to support displaced people who have pre-
incident factors in the Social Determinants of Vulnerability Framework, particularly low-to-no
income, older adults, people of color, and those who are socially isolated. The literature was
limited in research on the relationship between homelessness and other social characteristics of
vulnerability. Consequently, homelessness did not have the number of connections to other
social conditions to include it in the framework. Homelessness was related to post-incident
displacement and therefore warranted being included in the considerations for accessibility of
mass care services. After emergencies, cities should also consider reaching out to populations
that are associated with having a lack of access to post-incident services such as the people
without vehicles who rent and live in multi-story or multi-unit buildings, particularly those who
are low-income and people of color (Fothergill & Peek, 2004; U.S. Centers for Disease Control
and Prevention, 2010).
The critical transportation core capability “provides transportation (including
infrastructure access and accessible transportation services) for response priority objectives,
42
including the evacuation of people and animals, and the delivery of vital response personnel,
equipment, and services into the affected areas” (U.S. Department of Homeland Security, 2011,
p. 12). The original national capabilities, called the Target Capabilities List, combined critical
transportation and mass care services into a single capability referred to as citizen evacuation and
shelter-in-place. It provided a much more people-focused context:
“…the capability to prepare for, ensure communication of, and immediately execute the
safe and effective sheltering-in-place of an at-risk population (and companion animals), and/or
the organized and managed evacuation of the at-risk population (and companion animals) to
areas of safe refuge in response to a potentially or actually dangerous environment. In addition,
this capability involves the safe reentry of the population where feasible” (U.S. Department of
Homeland Security, 2007, p. 377).
There are significant logistical considerations with regard to sheltering and evacuating
people, particularly those who are most vulnerable. Estimates for the percentage of Americans
with a disability ranges from 19 to 30% with over 10 million people that have a vision disability
–blind, low vision, deaf/blind – and cannot see a map on television that shows them evacuation
routes (Federal Emergency Management Agency, 2011a). The ability to evacuate people
includes the need for public information and warning and mass care services capabilities.
Since most Americans live in urban areas, they need fewer cars, have better public
transit, can share cars, and accomplish more trips with walking (Addison, 2010). Additionally,
“[m]any physically and economically disadvantaged people depend on public transportation to
access to medical services and obtain healthy, affordable food” (Litman, 2010, p. 1). The lack of
vehicles and dependence on public transportation means there are more people who may need
support to evacuate from cities.
43
The Social Determinants of Vulnerability Framework does not include people without
vehicles. However, the condition of people without vehicles was related to social isolation,
limited access to post-incident services, low-to-no income, older adults, and people with
disabilities, all of which are part of the Social Determinants of Vulnerability Framework. The
Framework accounts for people without vehicles through the relationship with other social
conditions listed above.
Recovery. As a community begins to rebuild and create a new normal, capabilities that
have implications for socially vulnerable populations include health and social services, housing,
and economic recovery. Health and social services is the restoration and improvement of health
and social services networks to promote the resilience, independence, health (including
behavioral health), and well-being of the whole community (U.S. Department of Homeland
Security, 2011). This capability includes considerations for restoration of health and social
services based on at-risk individuals. FEMA attempted to define at-risk individuals, but only
included children, people with disabilities, limited English proficiency, and people with access
and functional needs, which was originally a reference to people with disabilities (Kailes &
Enders, 2007) that FEMA has adapted to encompass a larger but more nebulous range of socially
vulnerable populations. The challenge with the terminology of access and functional needs is that
it only focuses on the needs of some people after emergencies (there are no considerations for
low-income, people of color, socially isolated, among others) and the context is predominantly
about evacuation and sheltering.
Ideally, emergency plans exist for populations who need the most support and resources
to ensure they are safely evacuated or sheltered (Federal Emergency Management Agency,
2009). However, “…our country has yet to address the long-term affects disasters can have on
44
families and individuals who suffer through them” (Hansell, 2009, p. ). An example Hansell
provided was that after Hurricane Katrina, there was no plan in place to properly support the
return of Louisiana’s citizens to their homes or to provide for the needs of residents once they
arrived. As a result, thousands of the most vulnerable populations continue to suffer a long
recovery process.
The analysis that resulted in the Social Determinants of Vulnerability Framework
indicates that socially vulnerable populations have multiple obstacles to accessing post-incident
resources and benefits made available as part of the recovery process. These barriers including
being a renter, lack of access to a vehicle, being homeless, low-to-no income, older adults,
person of color (particularly Latino/Hispanic), and social isolation. The Framework provides a
straightforward and clear method for local government to identify people who are most
vulnerable to focus outreach and make better decisions about the provision of human services as
well as housing and economic recovery. This approach will help reduce the likelihood of
prohibited housing discrimination and inequity in economic stabilization of communities as
witnessed post-Katrina (Muñiz, 2006).
This research outlines social factors that result in a higher pre-incident allostatic load (the
sum of the physical and mental health impacts from stressful or traumatic events). Culturally
appropriate trauma response and mental health services should be made available to help
normalize their acute stress reactions and provide coping strategies to reduce the likelihood of
post-traumatic stress syndrome and other mental health illness. This is a post-incident mitigation
strategy that will improve their ability to deal with the cumulative stressors in their daily lives.
Relatedly, people should be provided with post-incident human services that are trauma
informed and help people address their needs associated with the emergency, and put people in a
45
position to improve their pre-disaster circumstances. The Social Determinants of Vulnerability
Framework can be used to focus recovery efforts to improve access to human services for all
members of the community.
Legal Compliance and Assistance
The exclusion, or at least the lack of inclusion, of socially vulnerable people does not
seem to be a deliberate action by local jurisdictions. It appears to be a lack of resources and
education about the responsibility of a city to develop, implement, and execute plans to mitigate,
respond, and recover in a manner that does not exclude people from the planning process and
accessing resources. The FEMA Civil Rights Program Director’s Policy issued in 2005 prohibits
organizations receiving funding from FEMA from denying the benefits of or participation in said
programs or activities on the basis of race, color, national origin, sex, religion, age, disability, or
economic status. However, it only explicitly focuses on disaster assistance functions in the
aftermath of an incident.
The plans, trainings, and exercises for response and recovery are based on the results of
the pre-incident mitigation capabilities, particularly risk and risk assessment, community
resilience, and long term vulnerability reduction capabilities. It would be beneficial to local
jurisdictions for FEMA to explain with more detailed guidance that pre-incident assessments and
planning for mitigation, response, and recovery are activities that have to be inclusive. Technical
assistance would be helpful for jurisdictions to assess their planning processes and the resultant
plans to develop a realistic set of corrective action steps to be more inclusive. The Social
Determinants of Vulnerability Framework at a minimum provides social conditions to consider
for inclusive planning, but more significantly, is a guide that municipalities can use to identify
the social circumstances and the corresponding neighborhoods within their jurisdiction with
46
which they can build stronger relationships and partnerships to facilitate more inclusive planning
processes and plans.
Social vulnerability encompasses the physical, psychological, social, and economic
considerations of pre-incident planning and post-incident impacts on people. These are the key
focus areas for the public health role in emergency management. It would be helpful if FEMA
coordinated with the U.S. Department of Health and Human Services, including the Center for
Disease Control and Preventions, to align language and guidance for public health and
emergency management collaboration. Currently, the Public Health Preparedness Capabilities:
National Standards for State and Local Planning provides the context for the public health role in
emergency management (U.S. Centers for Disease Control and Prevention (CDC), 2011).
However, it was developed based on the former Target Capabilities List (TCL), which FEMA
has since stopped using in favor of the core capabilities. A public health preparedness
practitioner has to do a crosswalk from the Public Health Preparedness Capabilities to the TCL
and then from the TCL to the core capabilities in order to align their terminology and work with
their emergency management colleagues. Similarly, the National Health Security Strategy
(NHSS) provides a comprehensive vision for protecting people’s health in an emergency based
on the nation’s goals for resilient communities and strong public health and healthcare systems
response and recovery systems (U.S. Department of Health and Human Services, 2009). The
NHSS is a comprehensive attempt to bring together the role of community resilience at the
individual, organizational, and government levels as well as emergency management and
considerations for at-risk populations. Unfortunately, the NHSS is also based on the Target
Capabilities List. The lack of alignment between guidance for emergency management and
47
public health is clearly an impediment for their ability to collaboratively improve inclusive
emergency planning.
Furthermore, partnerships between regional FEMA offices, the state, and cities should be
formalized to conduct social vulnerability analysis and inform whole community outreach and
planning strategies. This way, FEMA can work with cities to develop and implement strategies
with local agencies such as civic, community, faith-based, and private sector organizations, to
facilitate inclusive emergency planning. As part of the grant process for FEMA programs,
jurisdictions should have to conduct a social vulnerability analysis, assess existing plans based
on the results, and demonstrate the manner in which they are incrementally addressing the
identified corrective actions for community inclusiveness and engagement. The exclusion, or at
least the lack of inclusion, of socially vulnerable people does not seem to be a deliberate action
by local jurisdictions. It appears to be a lack of resources and education about the responsibility
of a city to develop, implement, and execute plans to mitigate, respond, and recover in a manner
that does not exclude people from the planning process and accessing resources. The FEMA
Civil Rights Program Director’s Policy issued in 2005 prohibits organizations receiving funding
from FEMA from denying the benefits of or participation in said programs or activities on the
basis of race, color, national origin, sex, religion, age, disability, or economic status. However, it
only explicitly focuses on disaster assistance functions in the aftermath of an incident.
The plans, trainings, and exercises for response and recovery are based on the results of
the pre-incident mitigation capabilities, particularly risk and risk assessment, community
resilience, and long term vulnerability reduction capabilities. It would be beneficial to local
jurisdictions for FEMA to explain with more detailed guidance that pre-incident assessments and
planning for mitigation, response, and recovery are activities that have to be inclusive. Technical
48
assistance would be helpful for jurisdictions to assess their planning processes and the resultant
plans to develop a realistic set of corrective action steps to be more inclusive. The Social
Determinants of Vulnerability Framework at a minimum provides social conditions to consider
for inclusive planning, but more significantly, is a guide that municipalities can use to identify
the social circumstances and the corresponding neighborhoods within their jurisdiction with
which they can build stronger relationships and partnerships to facilitate more inclusive planning
processes and plans.
Social vulnerability encompasses the physical, psychological, social, and economic
considerations of pre-incident planning and post-incident impacts on people. These are the key
focus areas for the public health role in emergency management. It would be helpful if FEMA
coordinated with the U.S. Department of Health and Human Services, including the Center for
Disease Control and Preventions, to align language and guidance for public health and
emergency management collaboration. Currently, the Public Health Preparedness Capabilities:
National Standards for State and Local Planning provides the context for the public health role in
emergency management (U.S. Centers for Disease Control and Prevention (CDC), 2011).
However, it was developed based on the former Target Capabilities List (TCL), which FEMA
has since stopped using in favor of the core capabilities. A public health preparedness
practitioner has to do a crosswalk from the Public Health Preparedness Capabilities to the TCL
and then from the TCL to the core capabilities in order to align their terminology and work with
their emergency management colleagues. Similarly, the National Health Security Strategy
(NHSS) provides a comprehensive vision for protecting people’s health in an emergency based
on the nation’s goals for resilient communities and strong public health and healthcare systems
response and recovery systems (U.S. Department of Health and Human Services, 2009). The
49
NHSS is a comprehensive attempt to bring together the role of community resilience at the
individual, organizational, and government levels as well as emergency management and
considerations for at-risk populations. Unfortunately, the NHSS is also based on the Target
Capabilities List. The lack of alignment between guidance for emergency management and
public health is clearly an impediment for their ability to collaboratively improve inclusive
emergency planning.
Furthermore, partnerships between regional FEMA offices, the state, and cities should be
formalized to conduct social vulnerability analysis and inform whole community outreach and
planning strategies. This way, FEMA can work with cities to develop and implement strategies
with local agencies such as civic, community, faith-based, and private sector organizations, to
facilitate inclusive emergency planning. As part of the grant process for FEMA programs,
jurisdictions should have to conduct a social vulnerability analysis, assess existing plans based
on the results, and demonstrate the manner in which they are incrementally addressing the
identified corrective actions for community inclusiveness and engagement.
Next Steps and Future Research
The Social Determinants of Vulnerability Framework is rooted in a literature-based
qualitative analysis that provides a theoretical framework, but is limited in its helpfulness to
explain the strength in the relationships between social conditions. Further, the literature is
limited to issue areas researchers found interesting or important enough to study. There are gaps
in the literature that impact the full understanding of the relationships between social factors of
vulnerability. Additionally, the social realities of the population in each city are unique. The
Framework provides a consistent baseline for all cities to start. There may be social conditions
that the planner and stakeholders are aware of based on their experience and knowledge of the
50
city. Additional categories can be added to a correlation analysis to determine if they are related
to the social determinants of vulnerability, particularly social isolation.
The next step in my research process is to collect data for the City of Boston at the census
tract level for each of the seven social conditions of the framework and analyze the correlation of
the social characteristics of vulnerability. This step will determine if the relationships between
social conditions the literature identified hold true at the local level and identify the strength of
the relationships. Geographic information systems will be used to visualize the distribution of the
social conditions in the framework to determine hotspots to focus emergency planning efforts.
Another component of the next phase of research is to explore the relationship of social isolation
with the other social characteristics in the framework using a regression analysis based on the
Boston data. In this research project, I found that social isolation had the most links to other
social characteristics among vulnerable populations. Social isolation is a pre-incident predictor of
mortality comparable to traditional clinical risk factors such as smoking and high blood pressure
(Pantell, Rehkopf, Jutte, Syme, Balmes, & Adler, 2013).
An unintended consequence of developing this framework using a link analysis approach
via TouchGraph software was that I created a tool that any city planner or researcher could use to
explore the relationships among the 63 social characteristics I started with and the associated
social conditions related or conduct additional research on social conditions to further inform the
relationships between the social conditions that increase vulnerability. My hope is that I will be
able to obtain funding to share the interactive version of this Framework through publishing the
full link analysis to the internet using ToughGraph Navigator Web.
51
Conclusion
Communities are faced with persistent, ongoing stressors that lead to poor health
outcomes. An emergency or disaster exposes vulnerable populations to intensified social
adversity that further compounds existing stressors. If emergency management plans and actions
are not community focused, jurisdictions will increase the poor health and social outcomes for
those most vulnerable thereby directly contributing to the long-term poor physical, mental, and
emotional health outcomes. Furthermore, if the resources to support those who need them most is
not accessible, then local government has broken one of the four key pillars of public
administration: social equity. The core components of social resilience are physical and
psychological health as well as the social and economic well-being (Chandra et al., 2011).
Everyone should have access to resources and decision making processes that improve these
aspects of health and well-being before, during, and after emergencies.
Social justice and human rights principles should guide the entire spectrum of emergency
planning from disaster risk management, including pre-disaster mitigation and community
resilience and preparedness measures, to recovery, including emergency relief and rehabilitation,
and reconstruction efforts (Brookings-Bern Project on Internal Displacement, 2008).
Incorporating the needs of socially vulnerable populations is recognition that they are part of the
community and therefore deserve the same planning considerations as anyone else.
I reviewed a purposive sampling of documents in order to build the Social Determinants
of Vulnerability Framework. This framework identifies the social conditions that are most
represented in the literature and are associated with higher likelihood of suffering physical
injuries or illness, psychological consequences, social disruption, and economic impact. The
framework provides a structure for cities to develop inclusive mitigation, response, and recovery
52
planning strategies while fostering social equity in public administration and reducing the
likelihood of successful litigation.
Invest Now or Pay Later
Local jurisdictions have an opportunity to take concrete actions to improve the resilience
of people in their cities. Litigation successfully brought against New York and Los Angeles
presents some of the financial and reputational costs of failure to include the social conditions of
people. Moreover, the cost of lives and unnecessary suffering for those most vulnerable is far too
high of a price to pay when the alternative is an investment of time and relationship development
with socially vulnerable people and the organizations that serve them. The return on investment
not only strengthens resilience to emergencies, but is consistent with improvement in other
community issues such as violence, health, and education. The factors that have to be considered
to make communities in cities safer, healthier, and more sustainable are the same protective
factors that make cities more resilient to disasters. Building stronger communities hinges on a
more sophisticated understanding of the interplay between the social conditions of people living
in our communities. The Social Determinants of Vulnerability Framework informs practical,
inclusive emergency planning that reduces the unnecessary suffering of people in American
cities.
53
References
Brookings-Bern Project on Internal Displacement. (2008). Human Rights and Natural Disasters:
Operational Guidelines and Field Manual on Human Rights Protection in Situations of
Natural Disaster. Washington, D.C.
Chandra, A., Acosta, J., Stern, S., Uscher-Pines, L., Williams, M. V., Yeung, D., . . . Meredith,
L. S. (2011). Building Community Relience to Disasters: A Way Forward to Enhance
National Health Security. Santa Monica: RAND Corporation.
Cutter, S. L., & Emrich, C. T. (2006). Moral Hazard, Social Catastrophe: The Changing Face of
Vulnerability along the Hurricane Coasts. Annals of the American Academy of Political
and Social Science, 604(1), 102-112.
Dickson, E., Baker, J. L., & Hoornweg, D. (2012). Urban Risk Assessments: Understanding
Disaster and Climate Risk in Cities: World Bank Publications.
Durant, T. J. (2011). The Utility of Vulnerability and Social Capital Theories in Studying the
Impact of Hurricane Katrina on the Elderly. Journal of Family Issues, 32(10), 1285-1302.
doi: 10.1177/0192513X11412491
Dwyer, A., Zoppou, C., Nielsen, O., Day, S., & Roberts, S. (2004). Quantifying Social
Vulnerability: A methodology for identifying those at risk to natural hazards. Geoscience
Australia Record, 2004/14.
Federal Emergency Management Agency. (2010). Guidance on Planning for Integration of
Functional Needs Support Services in General Population Shelters. Washington, D.C.
Federal Emergency Management Agency. (2011). National Disaster Recovery Framework.
Washington, D.C.: U.S. Department of Homeland Security.
54
Flanagan, B. E., Gregory, E. W., Hallisey, E. J., Heitgerd, J. L., & Lewis, B. (2011). A Social
Vulnerability Index for Disaster Management. Journal of Homeland Security and
Emergency Management: Vol. 8: Iss. 1, Article 3, 1-21.
Fothergill, A., & Peek, L. A. (2004). Poverty and disasters in the United States: A review of
recent sociological findings. Natural Hazards, 32(1), 89-110.
Galea, S., Freudenberg, N., & Vlahov, D. (2005). Cities and population health. Social Science &
Medicine, 60(5), 1017-1033. doi: http://dx.doi.org/10.1016/j.socscimed.2004.06.036
Ganzel, B. L., Morris, P. A., & Wethington, E. (2010). Allostasis and the human brain:
Integrating models of stress from the social and life sciences. Psychol Rev, 117(1), 134-
174. doi: 10.1037/a0017773
Greenberger, M. (2007). Preparing Vulnerable Populations for Disaster: Inner-city Emergency
Preparedness - Who Should Take the Lead. Journal of Health Care Law, 291-308.
Gustafsson, P. E., San Sebastian, M., Janlert, U., Theorell, T., Westerlund, H., & Hammarström,
A. (2014). Life-Course Accumulation of Neighborhood Disadvantage and Allostatic
Load: Empirical Integration of Three Social Determinants of Health Frameworks.
American Journal of Public Health, 104(5), 904-910.
Hansell, D. (2009). Disaster Case Management: Developing a Comprehensive National Program
Focused on Outcomes, written statement for Ad Hoc Subcommittee on Disaster
Recovery. Paper presented at the Committee on Homeland Security and Governmental
Affairs, US Senate, .
Isaranuwatchai, W., Coyte, P. C., McKenzie, K., & Noh, S. (2013). Impact of the 2004 Tsunami
on Self-Reported Physical Health in Thailand for the Subsequent 2 Years. American
Journal of Public Health, 103(11), 2063-2070.
55
Issenberg, S. (2012). The victory lab: The secret science of winning campaigns: Random House
LLC.
Kailes, J. I., & Enders, A. (2007). Moving Beyond "Special Needs". Journal of Disability Policy
Studies, 17(4), 230-237.
Klinenberg, E. (2002). Heat wave: a social autopsy of disaster in Chicago. Chicago: The
University of Chicago Press.
Knobel, C. (2013). Social network analysis as an augmentation of qualitative research. Paper
presented at the Collaboration Technologies and Systems (CTS), 2013 International
Conference.
Litman, T. (2010). Evaluating public transportation health benefits. Victoria, British Columbia,
Canada: Victoria Transport Policy Institute.
Logan, J. G., & Barksdale, D. J. (2008). Allostasis and allostatic load: expanding the discourse
on stress and cardiovascular disease. Journal of Clinical Nursing, 17(7b), 201-208. doi:
10.1111/j.1365-2702.2008.02347.x
Milligan & Company, L., ,. (2007). Transportation Equity in Emergencies: A Review of the
Practices of State Departments of Transportation, Metropolitan Planning Organizations,
and Transit Agencies in 20 Metropolitan Areas (Vol. 2014).
Mohaupt, S. (2009). Review Article: Resilience and Social Exclusion. Social Policy and Society,
8(1), 63-71. doi: 10.1017/S1474746408004594
Molin Valdes, H., Rego, A., Scott, J., & Aguayo, J. V. (2012). How to Make Cities More
Resilient: A Handbook for Local Government Leaders. In P. Bittner (Ed.). Geneva.
Morrow, B. H. (1999). Identifying and Mapping Community Vulnerability. Disasters, 23(1), 1-
18.
56
Muñiz, B. (2006). In the eye of the storm: how the government and private response to Hurricane
Katrina failed Latinos. Washington, DC.
Norman-Major, K. (2011). Balancing the Four Es; or Can We Achieve Equity for Social Equity
in Public Administration? Journal of Public Affairs Education, 17(2), 233-252.
Pantell, M., Rehkopf, D., Jutte, D., Syme, S. L., Balmes, J., & Adler, N. (2013). Social isolation:
a predictor of mortality comparable to traditional clinical risk factors. American Journal
of Public Health, 103(11), 2056-2062. doi: 10.2105/AJPH.2013.301261
Pantell, M., Rehkopf, D., Jutte, D., Syme, S. L., Balmes, J., & Alder, N. (2013). Social Isolation:
A Predictor of Mortality Comparable to Traditional Clinical Risk Factors. American
Journal of Public Health, 103(11), 2056-2062.
Paulison, R. D. (2005). Civil Rights Program. Washington, D.C.: U.S. Department of Homeland
Security.
Pelling, M. (2003). Vulnerability in Cities. London: Earthscan Publications.
Seeman, T. E. (1996). Social ties and health: The benefits of social integration. Annals of
epidemiology, 6(5), 442-451.
Sherry, J. M., Nishamarie, & Harkins, J. D., Anne Marie. (2011). Leveling the emergency
preparedness playing field. Journal of Emergency Management, 9(6), 11-16. doi:
10.5055/jem.2011.0075
Shi, L., & Stevens, G. D. (2010). Vulnerable populations in the United States (Vol. 23): John
Wiley & Sons.
U.S. Census Bureau (Producer). (2013, August 11). Growth in Urban Population Outpaces Rest
of Nation, Census Bureau Reports. Census.gov. Retrieved from
http://www.census.gov/newsroom/releases/archives/2010_census/cb12-50.html
57
U.S. Centers for Disease Control and Prevention. (2010). Public Health Workbook: To Define,
Locate, and Reach Special, Vulnerable, and At-Risk Populations in an Emergency.
Washington, D.C.
U.S. Centers for Disease Control and Prevention (CDC). (2011). Public Health Preparedness
Capabilities: National Standards for State and Local Planning. Atlanta.
U.S. Department of Health and Human Services. (2009). National Health Security Strategy of
the United States of America. Washington, D.C.
U.S. Department of Homeland Security. (2011). National Preparedness Goal, First Edition.
Washington, D.C.
U.S. Department of Homeland Security. (2013a). National Mitigation Framework. Washington,
D.C.
U.S. Department of Homeland Security. (2013b). National Response Framework (Second ed.).
Washington, D.C.
United Nations Office for Disaster Reduction. (2005). Hyogo Framework for Action 2005-2015:
Building the Resilience of Nations and Communities to Disasters. Geneva.
Wisner, B., & Handmer, J. (1999). Hazards, globalisation, and sustainability. Paper presented at
the Development in Practice.
58
Appendix A: Abridged Link Analysis Results
Timeframe Measure/Variable Primary Attribute Pre-incident Access Lack of Health Insurance
Access Lack of Public Transportation Access Lack of Vehicle Access Technology Access Lack of Citizenship/Legal Status Access Immigrants/Refugees Age Children Age Older Adults (65 and older) Disability Cognitive/Developmental Disability Physical/Mobility Disability Disability Sensory Disability People with Disabilities Disease/Illness Chronic and Acute Medical Illness Disease/Illness Psychological Illness Disease/Illness Alcohol Dependency Disease/Illness Drug Dependency Education Less than High School Diploma Family Composition Single Parent Family Composition Primary/Sole Caregiver Family Composition Unmarried/Single Gender, Gender Identification, and Sexual Orientation
Women Lesbian, Gay, Bisexual, and Transgender
Group Quarters Adult Correctional Facilities College/University Student Housing Juvenile Facilities Daycare Centers/Schools Nursing Homes
Household Composition Head of Household Living Alone
Housing Type Renters High-Rise Multi-Story/Unit Buildings
Income Low-Income Poverty Low-to-No Income Unemployed
59
Timeframe Measure/Variable Primary Attribute Homeless
Language/Literacy Limited English Proficiency Limited Literacy Proficiency
Living Conditions High Population Density No Open/Green Space Living spaces with fewer rooms High-Crime Areas
Race Asian Black Latino/Hispanic Native American People of Color
Social Connectedness Low Voter Turnout Low Political Engagement Social Isolation
Temporary Populations Tourists Commuters
Working Conditions
Outdoor Workers Responders
Post-Incident
Outcomes/Loss Access to Services Injury, Illness, or Death Loss of Business Loss of Employment Loss of Loved One Property Damage Displacement Domestic Violence
60
Code Code Title Code Explanation/Description Values Example TIME Timeframe Description of timeline of social
factors occurrence or significance Text Pre-Incident
MEAS Measure/Variable Measure of social vulnerability; major category usu. has multiple attributes
Text Age
ATTR Primary Attribute Specific attribute of social vulnerability; a single category that may have co-occurring categories
Text Children
ATTRSRC Attribute Source(s) Author and year of the source(s) for the social vulnerability attribute
Text Hajat et al., 2003; Knowlton et al., 2009; AAP, 2000; Doocy et al., 2013
ASCITE As Cited In Source being cited within a source, where applicable
Text Cooley, Moore, Heberger and Allen 2012
ASSCATTR Associated Attributes
Secondary attributes associated with the 'Primary Attribute'
Text Low-to-No Income; People of Color; Refugee or Immigrant; Limited English Proficiency; Institutionalized Settings; Homeless or Runaway; Chronic and Acute Medical Illness/Disease; Cognitive/Developmental; Physical Disabilities; Psychiatric Disorders; Home Alone; Daycare Centers/Schools
ASSCATTRSRC Associated Attributes Source(s)
Author and year of the source(s) for the associated attribute
Text Pfefferbaum and Shaw, 2013
61
Code Code Title Code Explanation/Description Values Example NOTES Notes Notes regarding the Primary
Attribute Text Institutionalized Settings (foster care homes,
halfway houses, shelters for domestic violence, and youth hostels); "Disasters can undermine the systems of safety that are in place to protect children, leaving them vulnerable to secondary stressors associated with violence, abuse, and opportunistic crimes."-Pfefferbaum and Shaw, 2013 Children are more vulnerable because they depend on others for care.-Shi and Stevens, 2010
62
Appendix C: Abridged Link Analysis Results
Figure 1 is the visualization of the most relevant part of the link analysis in the TouchGraph Navigator software, which includes pre-incident and post-incident factors. The pink boxes are the 63 pre-incident factors I started with and the purple or those that the literature indicated were associated with them. The post-incident outcomes have links to the large Post-Incident label and highlighted with a yellow box. The larger the size of the halo or circle, the larger the number of associations with other social factors.
63
Appendix D: Social Determinants of Vulnerability Framework
64
Paper 2: Application of the Social Determinants of Vulnerability Framework to the City of Boston
Background
A common cliché in emergency management is that all disasters are local. This
expression is particularly true when it comes to meeting the diverse needs of people living in
American cities. In fact, the national mitigation, response, and recovery frameworks that guide
emergency management in America each identify local governments as having the largest role
and responsibility. However, local governments have been slow to engage diverse communities
and incorporate their needs into mitigation, response, and recovery planning.
Civil rights statutes and federal guidance protect people in the United States from being
denied the benefits of, excluded from participation in, or subject to discrimination under
federally funded programs or activities on the basis of race, color, national origin, disability, age,
limited English proficiency, or economic status (Milligan & Company, 2007; Paulison, 2005;
U.S. Centers for Disease Control and Prevention, 2010). The lack of inclusion, or exclusion, of
socially vulnerable people does not seem to be a deliberate action by local jurisdictions. It
appears to be a lack of resources and education about the responsibility of a city to develop,
implement, and execute plans to mitigate, respond, and recover in a manner that has the effect of
excluding people from the planning process and accessing resources.
Many of the social factors in the Social Determinants of Vulnerability Framework (see
Appendix A) developed as part of this research includes all of the legally protected categories of
people. The Framework helps to minimize the likelihood that people’s rights are violated. This
Framework at a minimum, describes the relationship between pre-incident social conditions and
post-incident outcomes to be used for inclusive planning. More significantly, it is a guide that
municipalities can use to identify the social circumstances and the corresponding neighborhoods
65
within their jurisdiction with which they can build stronger relationships and partnerships to
facilitate a more inclusive planning processes.
When emergency management and public health preparedness considerations are not
inclusive, it is also a matter of homeland security and national health security (U.S. Department
of Health and Human Services, 2009; U.S. Department of Homeland Security, 2014). National
health security is defined as “…a state in which the Nation and its people are prepared for,
protected from, and resilient in the face of health threats or incidents with potentially negative
health consequences” (U.S. Department of Health and Human Services, 2009, p. 3). Building
resilience in communities is a primary goal of the National Health Security Strategy. The
strategy integrates considerations for at-risk populations throughout each of the capabilities for
national health security.
At least 50% of the United States can be considered socially vulnerable (Kailes &
Enders, 2007). Using a similar method as Kailes & Enders, the City of Boston has at least 36%
of the population that may be considered socially vulnerable (U.S. Census Bureau, 2012).
However, social vulnerability is not an issue that only affects those people. It consists of people
in our families, our children, our coworkers, and friends. Over two hundred thousand residents in
the City of Boston are affected.
This research focuses on developing a replicable approach to understanding the
complexity of social vulnerability in American cities for policy makers and emergency
management practitioners across all sectors of government and industry, particularly public
health emergency preparedness. Local emergency planners can use the Framework to: (1) map
pre-incident social factors and determine the high concentration areas; (2) understand how each
of the factors are related to other social conditions and characteristics; (3) develop outreach plans
66
to partner with those communities, (4) conduct an inclusive emergency planning process, and (5)
equitably execute the plans in the response and recovery phases based on the links between
social vulnerability and post-incident impacts. By using the Social Determinants of Vulnerability
Framework, local emergency management and public health preparedness planners can reduce
the likelihood of violating people’s rights and poor outcomes of disasters.
The Social Determinants of Vulnerability Framework was applied to the City of Boston
to determine if the relationships between social factors found in the literature apply to Boston
and explore the relationship between social isolation and the other social factors.
Social Determinants of Vulnerability Framework
The Social Determinants of Vulnerability Framework is rooted in a literature-based
qualitative analysis that provides a theoretical framework for identifying and incorporating the
diverse needs of people in American cities. There were seven pre-incident social factors that
seemed to be driving social vulnerability based on the number of links to other pre-incident
factors (See Table 1). These seven social conditions are directly or indirectly connected to six
post-incident outcomes that further increase social vulnerability.
Socially vulnerable people are disproportionately exposed to daily adversities. This
constant exposure to stressors deteriorates their physical and cognitive health resulting in a
comparatively higher allostatic load. Allostatic load is the sum of the body’s reactions to
stressful events. The types of exposures people experience in their lives before an emergency
that increase allostatic load include: exposure to illness; residential instability (including
displacement); death or illness of a close loved one; social isolation; limited opportunity to make
their own decisions; and exposure to threat or violence (Gustafsson et al., 2014). The results of
67
this research indicate that many of these same exposures are also driving increased vulnerability
after an emergency.
Table 1. Social Determinants of Vulnerability Framework: Pre-Incident Social Factors
Children
People with Disabilities
Older Adults
Chronic and Acute Medical Illness
Social Isolation
Low-To-No Income
People Of Color
Post-incident outcomes represent the types of consequences of an emergency or disaster
people may experience. There were a total of eight post-incident factors from the literature (See
Table 2). Six of the eight had at least one direct link to pre-incident social conditions. The lack of
access to post-incident services and displacement were related to the largest number of pre-
incident social characteristics. These post-incident impacts were mostly related to three pre-
incident social factors: social isolation, low-to-no income, and people of color. Low-to-no
income had the most links to post-incident outcomes.
Table 2. Social Determinants of Vulnerability Framework: Post-Incident Outcomes
Displacement
Access to Services
Injury, Illness, and Death
Property Damage or Loss
Domestic Violence
Loss of Employment
68
Purpose
The purpose of this research is to apply the Social Determinants of Vulnerability
Framework to the City of Boston to determine if the relationships between social conditions as
identified in the literature hold true at the local level. Additionally, this study identifies the
geographic distribution and the strength of the relationships between the social characteristics
that increase vulnerability. The research questions are: What are the correlations between the
social conditions of vulnerability in the City of Boston? What is the geographic distribution of
the correlated social conditions? Are the correlated social conditions predictors of social
isolation?
The social factors of vulnerability are closely related. Existing literature does not take
into account the manner in which social vulnerability factors are often compounded to produce
negative consequences before, during, and after emergencies. The sheer volume and unclear co-
occurrence of these factors becomes a practical challenge in identifying vulnerable populations
within a community and developing strategies to reduce their exposure to harmful public health
and safety consequences of emergencies.
Methods
I collected the data from the U.S. Census Bureau and SimplyMap at the census tract level
for the City of Boston. Specifically, I used the U.S. Census Bureau’s American Community
(ACS) 2008-2012 5-Year Estimates. They are the most reliable of the ACS estimates and have
data at the census tract level which was key to the success of this analysis. Data from SimplyMap
consisted of two sets from Easy Analytic Software, Inc. (EASI) discussed below. The data used
in this research were for the following variables: children, people with disabilities, older adults,
chronic and acute medical illness, social isolation, low-to-no income, people of color, limited
69
English proficiency, renters, less than high school education, women, and lack of vehicle. This
allowed for identification of relationships and patterns at the neighborhood and sub-
neighborhood level.
The social realities of the population in each city are unique. Therefore, I collected data
for the seven pre-incident social conditions and characteristics in the Social Determinants of
Vulnerability Framework as well as seven additional factors (limited English proficiency,
renters, less than high school, women, and lack of vehicle, institutionalized population in nursing
homes/skilled nursing facility, and populations in college/university student housing). The
addition of these social factors was based on my experience and knowledge in public health,
emergency management, and homeland security in the City of Boston.
Proxy Data
There were three data variables that I had to develop using proxy measures for, social
isolation, medical illness, and lack of vehicle. Social isolation was defined based on the
Berkman-Syme Social Network Index (SNI), which focused on marriage or partnership,
frequency of contact with friends and family, frequency of religious participation, and group
membership (Pantell et al., 2013). Data were additively combined for three variables. The first
variable was the sum of families that had either a female or male householder who had no spouse
present and children under 18. The second one is non-family households with persons older than
65 living alone. The final variable was people with no membership to: a fraternal order, body of
government, religious club, civic club, country club, business club, collectors club, union, school
or college board, church board, charitable organizations, or AARP. The limitation of this data is
that lack of membership may overlap with over adults living alone or single parent households.
This overlap may result in some overestimations in the number of people counted as socially
70
isolated. The source of this data is the Easy Analytic Software, Inc. (EASI) Mediamark Research
(MRI) annual study based on a national sample of 26,000 consumers used to develop a model at
the census tract level.
The second variable that is a proxy measure is medical illness which is based on EASI
data accessed through SimplyMap. EASI modeled the health statistics for the U.S. population
based upon age, sex, and race probabilities using U.S. Census Bureau data. The probabilities are
modeled against the census and current year and five year forecasts. Medical illness is the sum of
asthma in children, asthma in adults, heart disease, emphysema, bronchitis, cancer, diabetes,
kidney disease, and liver disease. A limitation is that these numbers may be over-counted as the
result of people potentially having more than one medical illness. Therefore, the analysis may
have greater numbers of people with medical illness within census tracts than actually present.
Overall, the analysis was based on the relationship between social factors. Therefore, the analysis
may not be impacted enough to distort the results.
The final variable, lack of a vehicle, was not intentionally a proxy. I pulled the data from
SimplyMap, which labeled it as Census data. However, upon further review of the metadata, it
was based on estimates calculated by the using 2012 ACS data from the U.S. Census Bureau. As
a result, the number of people without vehicles is smaller than they exist at the census tract level.
Mapping
Data collected from the U.S. Census Bureau American Communities Survey 2008-2012
5-year estimates and SimplyMap was joined with the 180 census tracts in the City of Boston by
GEOID10 in ArcGIS mapping software. ArcGIS was used to conduct a census tract based hot
spot analysis using the Getis-Ord Gi statistic to identify statistically significant spatial clusters of
hot spots and cold spots.
71
Each of the 180 census tracts were assigned to Boston neighborhoods as designated by
the Boston Redevelopment Authority planning districts. This allowed for a correlation analysis
with each social factor for the city as a whole and within each of Boston’s 16 neighborhoods.
The hot spots and the social factors highly correlated with social isolation were used to describe
the citywide and neighborhood social conditions and characteristics to inform emergency
management.
Correlation and Regression Analysis
I conducted a linear regression analysis at the city level with social isolation as the
dependent variable and the other socially vulnerable attributes independent variables. I also
conducted multiple correlation analyses using the Pearson correlation coefficient. The first
correlation analysis was based on citywide data at the census tract level. The correlations with a
Pearson r > 0.6 and P < 0.05 were included in the final list of social factors used to develop a
Social Determinants of Vulnerability Framework specifically for the City of Boston. The second
set of correlation analyses were conducted at the neighborhood level. Since the variables were
known to be related to one another, I held a higher threshold for correlation (r > 0.70, P < 0.05)
for the neighborhood analysis. The social characteristics and conditions that correlated with
social isolation represent the most socially vulnerable people in the community. This was a
guiding principle in highlighting the most vulnerable groups within each of the Boston
neighborhoods.
Findings and Results
Based on the Social Determinants of Vulnerability Framework, the relationships between
the seven social conditions and characteristics in the City of Boston become more complicated.
Overall, the model remained relatively unchanged. However, some of the relationships between
72
Table 3. Correlations for Socially Vulnerable Populations in the City of Boston Social Vulnerability All Correlations (Pearson r)
Social Isolation Low-to-No Income (0.685); Less than High School (0.654); Women (0.663); Renters (0.648); People with Disabilities (0.873); Children (0.803); Older Adults (0.736); Limited English Proficiency (0.820); People of Color (0.762); Medical Illness (0.754)
Limited English Proficiency*
Social Isolation (0.820); Less than High School (0.674); Women (0.619); Medical Illness (0.666); Children (0.649); People of Color (0.735); Renter (0.734); People with Disabilities(0.823); Low-to-No Income (0.957)
People with Disabilities
Social Isolation (0.873); Older Adults (0.642); Children (0.789); Low-to-No Income (0.722); Limited English Proficiency (0.823); Less than High School (0.729); People of Color (0.803); Medical Illness (0.629)
People of Color Social Isolation (0.762); People with Disabilities (0.803); Children (0.801); Low-to-No Income (0.721); Limited English Proficiency (0.735); Less than High School (0.782)
Medical Illness Social Isolation (0.754); People with Disabilities (0.629); Older Adults (0.676); Limited English Proficiency (0.666); Women (0.967); Renters (0.706)
Renters* Social Isolation (0.648); Women (0.694); Low-to-No Income (0.701); Limited English Proficiency (0.734); Medical Illness (0.706); Lack of Vehicle (0.896)
Low-to-No Income Social Isolation (0.685); Less than High School (0.641); People with Disabilities (0.722); Limited English Proficiency (0.957); People of Color (0.721); Renters (0.701)
Children Social Isolation (0.803); Limited English Proficiency (0.649); People of Color (0.801); People with Disabilities (0.789); Less than High School (0.725)
Less Than High School*
Social Isolation (0.820); Low-to-No Income (0.641); Limited English Proficiency (0.957); People with Disabilities (0.729); Children (0.725)
Women* Social Isolation (0.663); Limited English Proficiency (0.619); Renter (0.694); Medical Illness (0.967)
Older Adults Social Isolation (0.736); Medical Illness (0.676); Disability (0.642)
Lack of Vehicle* Renters (0.896)
Institutionalized Population in Nursing
Homes/Skilled Nursing Facility*
N/A
College/University Student Housing*
N/A
*Added to the correlation analysis of social factors for the Boston Social Determinants of Vulnerability Framework. N/A: these social factors did not have a significant correlation with any of the Social Determinants of Vulnerability or each other.
73
the social conditions and characteristics in the framework are not as interrelated in Boston as the
literature suggests. For example, older adults and low-to-no income were not as correlated with
other social characteristics. Also, children were not correlated with medical illness or low-to-no
income.
The variables added on the basis of my public health and emergency management
experience and knowledge of Boston further complicated the model. Social isolation remained a
key variable and actually became even more significant in its direct relationship with of social
factors. As seen in Table 3, it was correlated with all attributes directly except people without
vehicles. However, people without vehicles were related to renters, which was directly
associated with social isolation.
Based on these results, I was able to modify the original Social Determinants of
Vulnerability Framework into a new framework specific to Boston. The original Social
Determinants of Vulnerability Framework, the Boston Social Determinants of Vulnerability
Framework and the expanded pre-incident Framework for Boston can be found in Appendix A.
Citywide Geographic Concentration of Social Determinants of Vulnerability Factors
People who have multiple social factors of vulnerability are likely to be more exposed to
negative post-incident outcomes than those who do not (Morrow, 1999). The mapping of each
social factor of the Boston Social Determinants of Vulnerability Framework is most helpful
when paired with the results from the citywide correlation analysis used to develop the
Framework. The goal is to understand the relationships between the social factors of
vulnerability at the neighborhood level. For each map, emergency planners can begin to consider
the other social factors people may be facing in those areas in the City of Boston.
74
The original Social Determinants of Vulnerability Framework indicated that post-incident
outcomes (lack of access to post-incident services; displacement; injury, illness, and death; loss
of employment; property damage; and domestic violence) were related to three social factors:
social isolation, low-to-no income, and people of color. Mattapan, Roxbury, and South
Dorchester had multiple census tracts with statistically significant concentrations of people who
were socially isolated, low-to-no income, and people of color. Therefore, these three
neighborhoods are most likely to have the largest exposure to post-incident impacts. The maps
from the hot spot analysis can be found in Appendix B.
Social Isolation and Social Vulnerability
The Social Determinants of Vulnerability Framework is based on a link analysis of data
from the literature to show the most connected social factors in vulnerability. Social isolation had
the most links to pre-incident social factors. The regression analysis confirmed the significance
of the relationship between the social isolation and the social factors in the Social Determinants
of Vulnerability Framework. As seen in Table 4, the model explained over 95% of the variation
in social isolation. Eight attributes explained social isolation the best: people with disabilities,
children, older adults, low-to-no income, less than high school education, people of color,
women, and renters. Further, people with limited English proficiency were highly correlated with
all eight social factors.
Social isolation has been validated as being driven by social vulnerability via link
analysis based on the literature, correlation analysis (citywide and within neighborhoods), and
regression analysis further solidifying the strong relationship between social isolation and social
vulnerability. For this reason, social isolation was used as the primary social factor to examine
vulnerability at the neighborhood level.
75
Social Determinants of Vulnerability within Boston Neighborhoods
The results suggest that one of the most socially vulnerable populations across the city is
women who do not have a vehicle, rent, and have a medical illness. In 12 out of 16
neighborhoods, lack of vehicle, medical illness, renter, and women were strongly correlated with
social isolation (r > 0.7 in every case, with most being r > 0.8). Additionally, these four variables
were clustered together in eight neighborhoods: Back Bay/Beacon Hill, Downtown, East Boston,
Hyde Park, Jamaica Plain, Mattapan, South Dorchester, and South End. Some of the clusters
were correlated with social isolation as well as other social factors, which are described in the
Table 4: Model of Social Isolation in Boston, MA Model Summary
Model R R Square Adjusted R
Square Std. Error of the
Estimate 1 .975a .951 .949 59.837
a. Predictors: (Constant), OCC_RENTER, TotChild, OlderAdult, LessThanHS, Women, Low_to_No, POC2, NoVehicle, TotDis, MedIllnes
Coefficientsa
Model Unstandardized Coefficients
Standardized Coefficients
t Sig. B Std. Error Beta 1 (Constant) 16.534 10.937 1.512 .132
TotDis .164 .042 .167 3.872 .000
TotChild .247 .021 .428 11.560 .000
OlderAdult .271 .036 .249 7.608 .000
Low_to_No -.050 .015 -.132 -3.379 .001
LessThanHS -.084 .027 -.100 -3.127 .002
POC2 .040 .008 .215 5.217 .000
Women -.065 .031 -.181 -2.089 .038
MedIllnes .055 .044 .127 1.251 .213
NoVehicle -.028 .029 -.041 -.965 .336
OCC_RENTER .241 .028 .484 8.483 .000
a. Dependent Variable: SocIsol
76
neighborhood summary below. In East Boston, Hyde Park, Mattapan, and Roxbury, social
isolation was correlated with all of the social factors in the Social Determinants of Vulnerability
Framework. The circumstances of socially vulnerable populations in these neighborhoods were
very closely related to one another and, therefore, they have higher vulnerability, as shown in
Appendix C.
The following provides an overview of the most socially vulnerable populations in each
of the neighborhoods. The analysis is based on the correlation analysis showing an existence of a
relationship between social isolation and each of the other social factors as described in Table 5
(r > 0.70, P < 0.05) and the social factors that had a high likelihood of being related to many
other social factors. The other consideration was for the geographic concentrations of socially
vulnerable people in those neighborhoods based on the hot spot analysis. The breakdown of
population for each neighborhood and associated social determinants of vulnerability are listed in
Appendix D.
77
Allston/Brighton. In Allston/Brighton, the most vulnerable population were older adults
who have a disability (r = 0.821, P < 0.01); and people with disabilities and limited English
proficiency who rent (r ≥ 0.731, P < 0.01). There were four census tracts (4.01, 5.03, 5.04, 6.02)
where the following social factors were concentrated: lack of vehicles, limited English
proficiency, medical illness, renters, low-to-no income, and women. The Boston Housing
Authority has four developments for people with low-income in those areas: John J. Carroll, for
people who are older adults or have a disability; Washington Street, for families and people who
are elderly or disabled; Commonwealth, predominantly for families with some units for people
Table 5. Neighborhood-Level Social Factors that Correlate with Social Isolation
Community
Soc
ial
Isol
atio
n
Dis
abil
itie
s
Ch
ild
ren
Old
er
Ad
ult
s
Low
-to-
No
Inco
me
LE
P
Les
s th
an
Hig
h
Sch
ool
Peo
ple
of
Col
or
Wom
en
Ren
ter
Med
ical
Il
lnes
s
Lac
k o
f V
ehic
le
Allston/Brighton 0.933 .489 .846 .685 .851 .684 .461 .532 .828 .669 .683
Back Bay/Beacon Hill
0.747 .791 .808 .654 .94 .405 0.871 .975 .942 .979 .899
Charlestown .746 .627 .442 .805 .854 .875 .811 .873 .969 .860 .956
Downtown .914 .832 .952 .656 .837 .671 .687 .868 .931 .966 .873
East Boston .925 .974 .843 .918 .954 .776 .887 .982 .982 .974 .929
Fenway/Kenmore .760 .444 .738 .701 .808 .595 .262 -.027 .922 .305 .940
Harbor Islands
Hyde Park .952 .864 .988 .971 .986 .886 .971 .987 .943 .988 .885
Jamaica Plain .918 .825 .680 .886 .905 .769 .863 .903 .892 .893 .866
Mattapan .918 .792 .768 .781 .868 .774 .977 .947 .986 .923 .968
North Dorchester .960 .825 .893 .233 .425 .770 .534 .746 .394 .836 .689
Roslindale .845 .913 .664 .813 .939 .350 .741 .954 .935 .931 .859
Roxbury .931 .916 .853 .871 .901 .720 .934 .802 .973 .771 .891
South Boston .725 .649 .805 .653 .823 .735 .473 .849 .943 .817 .877
South Dorchester .859 .815 .704 .375 .551 .702 .623 .969 .979 .963 .871
South End .556 .730 .535 .667 .729 .473 .698 .878 .901 .833 .854
West Roxbury .516 .789 -.004 0.91 .972 .745 0.918 .597 .956 .504 .582
r > 0.70
BOLD Hot Spot
Note: Detailed correlation matrices can be found in Appendix C.
78
who are elderly or disabled; and Patricia White, for people who are elderly or disabled (Boston
Housing Authority, 2014).
Back Bay/Beacon Hill. The most socially vulnerable people in Back Bay/Beacon Hill are
females of color who have limited English proficiency, rent, have a medical illness, and do not
have a vehicle (r ≥ 0.782, P < 0.01). This neighborhood had the smallest percent of people of
color as well as those with limited English proficiency compared to other neighborhoods. This
indicates that this cluster of social factors likely exists in smaller numbers.
Charlestown. Charlestown’s most vulnerable population were people of color with low-
to-no income, limited English proficiency, have less than a high school education, and who do
not have a vehicle (r ≥ 0.872, P < 0.01). Charlestown had the third smallest percentage of people
of color (24.22%) which is indicative that the number of people who may be most vulnerable is
relatively smaller.
Downtown. The most socially vulnerable population in Downtown Boston was older
women who had disabilities, limited English proficiency, rented, had a medical illness, no
vehicle, and children in the household (r ≥ 0.765, P < 0.01). It should be noted that the
Downtown area planning district includes Chinatown, West End, North End, and the Financial
District. It is a mixture of residential and commercial and has the third largest Chinese
neighborhood in the country (City of Boston, 2014). Downtown had the second largest
percentage of older adults, renters, medical illness, and renters.
East Boston. Social isolation in East Boston was highly correlated with all of the social
characteristics. The most socially vulnerable households in East Boston were likely to include all
or some combination of the following: having disabilities, households with children, low-to-no
income, limited English Proficiency, women, renters, medical illness, and no vehicle (r ≥ 0.703,
79
P < 0.01). Census tract 501.01 had a high concentration of people of color, with limited English
proficiency, disabilities, low-to-no income, children, and less than a high school education. This
census tract is in the Eagle Hill area where East Boston High School is located. Overall, East
Boston had the largest percentage of people without a high school education.
Fenway/Kenmore. The Fenway/Kenmore Square neighborhood had two groups of
people who were most socially vulnerable. First, low-income households that rent, do not have a
vehicle, and have limited English proficiency (r ≥ 0.764, P < 0.01). The second group is older
adults with a disability and limited English proficiency (r ≥ 0.73, P < 0.01). Fenway had the
smallest percentages of social isolation, children, and older adults. Additionally, it had the
second smallest percentages of people with disabilities and without a high school education. The
existence of smaller numbers of certain vulnerable groups suggests the second cluster may be a
relatively small group of people.
Hyde Park. In Hyde Park, social isolation was correlated with all of the social
determinants of vulnerability as well as each other (r ≥ 0.74, P < 0.01). Hyde Park had the third
largest percent of older adults as well as multiple areas with high concentrations of older adults.
Additionally, over 70% of the population was people of color who are concentrated within
multiple areas of the neighborhood. Like Fenway/Kenmore, there are populations with socially
vulnerable characteristics that occur in small numbers in Hyde Park. Interestingly, Hyde Park
had the smallest percentages of people who rent or have a medical illness and the second
smallest percent of people who do not have a vehicle.
Jamaica Plain. The most vulnerable population in Jamaica Plain was people with limited
English Proficiency based on their correlation with social isolation and all other social factors (r
≥ 0.704, P < 0.01). More than one third of the population in Jamaica Plain is low-to-no income
80
and over 40% have limited English proficiency. Jamaica Plain did not have any hot spots,
however, there was a census tract (1203.01) that was predominantly in Roxbury near the
Bromley-Health Housing Development (Boston Housing Authority, 2014) with hot spots for
low-to-no income and limited English proficiency.
Mattapan. Like Hyde Park, Roxbury and East Boston, social isolation in Mattapan is
correlated with all of the social determinants of vulnerability. People of color who rented and had
a disability are the most vulnerable populations in this neighborhood. These populations were
correlated with all other social factors and each other (r ≥ 0.712, P < 0.05). Mattapan had two
groups that had the second largest percentage of people compared to other neighborhoods:
people who were socially isolated and people with limited English proficiency. Additionally, this
neighborhood had the largest percentage of people with disabilities, children, and people of color
(over 95%).
North Dorchester. The most vulnerable populations in North Dorchester were
households with children, older adults, people with disabilities, individuals with less than a high
school education, and medical illness. All of these factors are correlated with social isolation and
each another (r ≥ 0.717, P < 0.05). This neighborhood had the second largest percentage of the
population with low-to-no income and less than a high school education.
Roslindale. Roslindale has two groups of people who were particularly vulnerable. The
first group is households with children, people with disabilities, low-to-no income, and limited
English proficiency (r ≥ 0.723, P < 0.05). The second group is women who rent and have a
medical illness (r ≥ 0.796, P < 0.01). Renters existed in smaller numbers in Roslindale than most
areas of the city. Three social factors have their own statistically significant geographic clusters:
people of color, older adults, and one census tract for children.
81
Roxbury. Like East Boston, Hyde Park, and Mattapan, social isolation in Roxbury was
correlated with all of the social factors of vulnerability. There were two distinct groups of people
who are most socially vulnerable: people of color who rented, were low-to-no income, had
children, had limited English proficiency, and had a disability (r ≥ 0.832, P < 0.01) and women
of color who rent and have a medical illness (r ≥ 0.823, P < 0.01). Census tract 9803 in the
southern most areas of Roxbury appeared to be a highly concentrated area of several social
factors of vulnerability. However, it was one of the two census tracts that includes the Franklin
Park Zoo and only had 338 people (most census tracts have more than ten times as many people).
The high concentrations are likely the result of the Lemuel Shattuck Hospital, which also has a
homeless shelter that some people may have claimed as their address. Census tract 814 that is
partially in Jamaica Plain and mostly in Roxbury was a geographic concentration for people who
are low-to-no income, people of color, and limited English proficient. This census tract includes
Roxbury Crossing, Roxbury Community College, the sub-neighborhoods Fort Hill and Highland
Park and extends to Washington Street in Roxbury near Malcom X/Washington Park. The
Boston Housing Authority owns a property in the Highland Park area and there are a few
subsidized MassHousing properties (Boston Housing Authority, 2014; MassHousing, 2014).
Roxbury had the largest percentage of people who were socially isolated, had limited English
proficiency, and low-to-no income. Additionally, it had the second largest percentage of people
with disabilities, children, and people of color (89.55%). Roxbury had the fourth smallest
percentage of older adults in the community.
South Boston. In South Boston, there were three distinct groups of socially vulnerable
populations: people with disabilities who had limited English proficiency and less than a high
school education (r ≥ 0.789, P < 0.01); female older adults with a medical illness that rent (r ≥
82
0.771, P < 0.01); and people who have limited English proficiency, less than a high school
education, and no vehicle (r ≥ 0.783, P < 0.01). Socially vulnerable groups occur in smaller
numbers in South Boston compared to other areas of the city. South Boston had the second
smallest population of people of color and relatively smaller percentages of people with
disabilities, children, older adults, low-to-no income, limited English proficiency, and people
with less than a high school education.
South Dorchester. The most vulnerable populations in South Dorchester were women
who lack a vehicle, rent, have a medical illness, children, and a disability (r ≥ 0.726, P < 0.01)
and women with less than a high school education, a disability, and children (r ≥ 0.710, P <
0.01). South Dorchester had the third largest percentages of people with disabilities, children,
less than a high school education, and people of color. This indicates that socially vulnerable
populations will exist in greater numbers in this neighborhood. Additionally, South Dorchester
had the largest number of hotspots compared to the other 15 neighborhoods.
South End. In the South End, there were two groups that were most socially vulnerable.
The first group was households with children and limited English proficiency (r ≥ 0.747, P <
0.05). The second group was women who rent, have a medical illness, and no vehicle (r ≥ 0.873,
P < 0.01). The South End has the third largest percentage of social isolation, low-to-no income,
limited English proficiency, renters, medical illness, and lack of a vehicle. More than half the
neighborhood was people of color. Census tract 711.01 within the South End had hot spots for
lack of a vehicle, low-to-no income, and people with disabilities. This area includes the Boston
Medical Center and Boston University Medical campuses, the Miranda-Creamer Building, the
residential towers owned by the Boston Public Health Commission, and several MassHousing
83
financed or subsidized housing for families, the elderly, and people with disabilities (Assessing
Department, 2014; MassHousing, 2014; Rosso, 2012).
West Roxbury. Socially vulnerable populations in West Roxbury are low-to-no income
households that rent, do not have a high school education, have limited English proficiency, and
are people of color (r = 0.764, P < 0.05). West Roxbury has the largest percentage of older
adults, however, this was not correlated with social isolation. Although West Roxbury has
groups of socially vulnerable populations, they do not exist in large numbers. The second
smallest percentage of people with limited English proficiency; the lowest percent of people
without a vehicle or who rent; and the third smallest percentages of people with less than a high
school education and low-to-no income.
Analysis and Recommendations
The reality of emergencies and traumatic incidents is that they often times may affect an
area of a neighborhood and not the entire city. The Social Determinants of Vulnerability
Framework provides an evidence-based, practical approach to building a more resilient Boston.
The Framework has made it possible to develop an outline of the neighborhoods, areas within
neighborhoods, and factors of vulnerability to focus mitigation, response, and recovery efforts.
This Framework also helps to be more informed in supporting the community during smaller
scale emergencies. Finally, the Framework can help to improve community resilience therefore
strengthening our national health security (U.S. Department of Health and Human Services,
2009).
Mitigation
Mitigation is intended to minimize the risks associated with threats and hazards (U.S.
Department of Homeland Security, 2013a). The most applicable mitigation capabilities to this
84
research are risk and disaster resilience assessment, community resilience, and long term
vulnerability reduction. Many plans do not identify the people most at-risk for poor outcomes
after emergencies. The plans that do include them, merely list them without a clear set of
mitigation actions that would be taken to reduce their risk.
Comprehensive mitigation strategies include incorporating social vulnerability into risk
assessments, which is the most foundational step in emergency management. The exclusion
from risk assessments increases the likelihood that mitigation, response, and recovery plans will
not account for their needs and potentially violates their civil rights.
Risk reduction strategies include community preparedness with a focus on building
resilience. The core components of social resilience are physical and psychological health as well
as the social and economic well-being (Chandra et al., 2011). Cities can increase social resilience
by working to reduce social isolation. For Boston, Social Isolation in East Boston, Hyde Park,
Mattapan, and Roxbury was correlated with all social factors of vulnerability. Connecting the
people in these neighborhoods to each other, resources in the community, and city agencies can
reduce social isolation. Additionally, Mattapan, Roxbury, and South Dorchester had
concentrations of vulnerability in specific census tracts associated with social isolation, low-to-
no income, and people of color, which are associated with more server poor outcomes after
emergencies. This information, along with the associated social factors for each neighborhood
provide guidance to Boston leadership for targeting community resilience efforts to best meet the
needs of its residents. Programs that increase these aspects of resilience can be focused in
neighborhoods, or areas within neighborhoods, with high levels of social vulnerability.
Mitigation strategies for socially vulnerable populations should include reducing social
isolation, which is the lack of social connectedness or social capital. This goes beyond the
85
individual’s close social networks to include community-level networks and institutions, both
public and private (Hawkins & Maurer, 2009). City agencies should develop relationships with
the organizations that serve these communities and populations (U.S. Department of Health and
Human Services, 2009). Two logical places to begin are the Boston Housing Authority and
MassHousing since they house residents within areas that have higher risk of vulnerability.
The BostonSocial Determinants of Vulnerability Framework can also inform the manner
in which preparedness materials are made accessible to those with limited English proficiency,
older adults, and people with disabilities (particularly sensory disabilities such as those with
limited vision and hearing). Additionally, resilience messaging can be provided to established
civic leaders trusted by the community, as trust is a necessary factor in building social capital
between individuals, the community, and institutions that serve them (Durant, 2011). These same
principles apply to communications and outreach associated with response and recovery.
Response
In this analysis, 16% of the population in Boston lack access to a vehicle, although the
U.S. Census Bureau estimates are as high as 36% (U.S. Census Bureau, 2012). Citywide, the
lack of a vehicle was correlated with renters. Within neighborhoods, the lack of a vehicle was
often closely related to social isolation, women, medical illness, and renters. These are key
considerations for evacuating and sheltering people before or after an emergency. They can
inform the types, amount, and locations for the deployment of necessary transportation assets. In
South Dorchester, people without vehicles also had children and a disability. Therefore, any
transportation provided for people with disabilities in that neighborhood may need to consider
that there may also be children to transport. In Hyde Park, however, transportation may not be
needed to the same degree because 95% of the population in this neighborhood owns a vehicle.
86
The social determinants of vulnerability analysis can also inform the potential needs of
people who may appear in shelters as well as the location of shelters and resources necessary to
transport people without vehicles to these locations. A shelter in Roxbury would need to be able
to accommodate a high number of children, people who may not speak English, have disabilities
or a medical illness. People working in these shelters would need to be knowledgeable about and
respectful of the range of cultures because just over half the people who live there have limited
English proficiency.
Recovery
After emergencies, Boston can leverage the Social Determinants of Vulnerability
Framework analysis to identify populations that are associated with having a lack of access to
post-incident services, such as the people without vehicles who rent and live in multi-story or
multi-unit buildings, particularly those who are low-income and people of color (Fothergill &
Peek, 2004; U.S. Centers for Disease Control and Prevention, 2010). Based on the literature,
exposure to post-incident impacts such as lack of access to post-incident services; displacement;
injury, illness, and death; loss of employment; property damage; and domestic violence are
associated with social isolation, low-to-no income, and people of color. For Boston, this means
that Mattapan, Roxbury, and South Dorchester are likely to have higher exposure to post-
incident impacts because these neighborhoods have multiple census tracts with hot spots for
social isolation, low-to-no income, and people of color.
Local government needs to understand the social factors of vulnerability at the
neighborhood level. This knowledge can help cities identify community and faith-based
organizations, city agencies, and other partners. The most effective partners would be those that
provided services to the affected community before an emergency or that have resources that
87
meet their needs after an emergency. Developing partnerships will also help create effective
post-disaster strategies to reach individuals in the community to facilitate access to post-incident
recovery resources.
Based on the Social Determinants of Vulnerability Framework, those living with low-to-
no income are at the highest risk for negative post-incident outcomes. Blaikie et al. (1999) noted
that low-income households have insufficient financial reserves for purchasing supplies in
anticipation of an event or for buying services and materials in the aftermath of one. “The
impact is likely to affect them disproportionately, including higher mortality rate” (as cited in
Morrow, 1999, p. 3). Their economic and material losses, while relatively less compared to other
economic groups, can be devastating because the loss is larger proportional to their total assets
(Morrow, 1999). Having low-to-no income in Boston was associated with social isolation, less
than high school education, people with disabilities, limited English proficiency, people of color,
and renters. Nine neighborhoods had hot spots for low-to-no income, Allston/Brighton, East
Boston, Fenway/Kenmore Square, Jamaica Plain (based on a census track predominantly in
Roxbury), Mattapan, North Dorchester, Roxbury, and South Dorchester. Allston/Brighton,
Roxbury, and South Dorchester had the most hot spots for low-to-no income.
This research outlines social factors that result in greater physical and mental health
impacts from stressful or traumatic events. Culturally appropriate trauma response and mental
health services should be made available to help normalize their acute stress reactions and
provide coping strategies to reduce the likelihood of post-traumatic stress syndrome and other
mental health illness. This is a post-incident mitigation strategy that will improve their ability to
deal with the cumulative stressors in their daily lives. People should be provided with post-
88
incident human services that are trauma informed and help people address their needs associated
with the emergency, and put people in a position to improve their pre-disaster circumstances.
The Social Determinants of Vulnerability Framework can be used to focus recovery
efforts to improve access to human services for all members of the community. A disaster
recovery center in the South End would need to have people who speak multiple languages and
people to help with medical needs. This includes connecting people with the healthcare system to
see a doctor that can help them manage their chronic or acute illnesses. This is an example of a
basic action that can help improve the long-term quality of life of people and reduce the post-
incident outcomes.
Socially vulnerable populations can maintain independence in their daily lives after an
emergency if they have access to the proper resources. This includes social and physical
infrastructure, such as health and human services, public transportation, and functioning utility
networks such as water, electricity and telecommunications (Dwyer, Zoppou, Nielsen, Day, &
Roberts, 2004). The quicker these at-risk groups are connected to supportive services after
emergencies, the more they will be able to maintain independence. Connecting the most
vulnerable populations to resources also reduces the post-incident cumulative stressors that can
further compound negative public health outcomes.
Legal Compliance and Assistance
Boston leadership can leverage this research to assess Boston’s emergency plans for
accounting for these populations, which are protected by civil rights statutes. They have a legally
protected right to be involved in the planning process, included in the emergency management
plans, and have access to post-incident services provide by local, state, federal, and
nongovernmental organizations. This reality demands that FEMA provide clearer guidance and
89
support for cities in America to build resilience in people in addition to the environment and the
infrastructure that supports them. FEMA can further support resilient communities by
incorporating the Social Determinants of Vulnerability Framework analysis into their risk
assessment for the Urban Area Security Initiative. This effort provides grant funding to many of
the larger American cities and encouraging cities to conduct this analysis for their jurisdictions.
The logical next step after the analysis, as recommended for Boston, is to assess emergency
plans for alignment with the needs of their communities.
A good model for providing guidance from a federal agency to partners is a Federal
Transit Administration’s report on civil rights and emergency preparedness (Milligan &
Company, 2007). This report provides guidance on practices that help transit authorities
understand the populations in their communities, their civil rights, and provide considerations for
evacuation. The Social Determinants of Vulnerability Framework can be used as a tool to update
the report. Additionally, FEMA and HHS can partner with the Federal Transit Administration to
develop similar guidance for American cities.
The U.S. Department of Homeland Security and the U.S. Department of Health and
Human Services can use the Social Determinants of Vulnerability Framework to partner with
cities to develop a similar report that also includes evidence-informed recommendations for
community engagement and inclusive planning that will be tied to funding. This requires that
emergency management and public health work closely together to develop a roadmap for how
they are going to increase their cities’ resilience.
Future Research
This research shows that the Social Determinants of Vulnerability Framework can
help to better understand the relationship between social factors of vulnerability in cities. This is
90
the first study of social vulnerability using this particular approach. The research was limited in
scope. However, I would like to continue this research to further explore each neighborhood and
better define the specific mitigation, response, and recovery actions that can be taken within the
neighborhoods of Boston.
The Boston Redevelopment Authority planning districts combine some neighborhoods
and separates Dorchester. The Boston Redevelopment Authority has a database which aligns the
planning districts with some of the smaller neighborhoods, such as Mission Hill, Longwood
Medical Area, North End, and West End (although it still does not have Chinatown or other well-
known sub-neighborhoods such as Grove Hall, Upham’s Corner, Fields Corner). I intend to
repeat the process for the City of Boston based on the smaller geographic units to determine if
there are any further lessons that can be learned about vulnerable populations in each area.
In the next iterations of this research, I will include the actual U.S. Census Bureau data
for people without vehicles and conduct a hot spot analysis for each of the most vulnerable
groups identified in the neighborhood correlation analyses to locate their concentrations at the
neighborhood level. Additionally, I would like to use the results to identify the organizations that
interact with the most vulnerable populations.
Conclusion
History has shown that socially vulnerable populations are disproportionately impacted
by disasters. These groups experience disproportionate suffering, particularly from public health
and safety impacts such as injuries, death, and a decreased likelihood of recovery (Cutter &
Emrich, 2006; Flanagan, Gregory, Hallisey, Heitgerd, & Lewis, 2011; Fothergill, Maestas, &
Darlington, 1999). Their social circumstances are complicated. However, current emergency
plans and risk assessments primarily focus on physical and environmental infrastructure (Mileti,
91
1999). These plans and assessments lack analysis of the social conditions and characteristics of
vulnerable individuals who need to be incorporated into mitigation, response, and recovery
plans.
In emergency management, it is easy to become distracted by the gadgets, consultants,
and the next new technology. It is important that local government does not lose sight of the
fundamental purpose of emergency management: to protect people. We have consistently
witnessed the devastating impacts of emergencies and large scale disasters on socially vulnerable
Americans.
Identifying and understanding the concentrations of specific segments of socially
vulnerable populations informs risk assessment and focuses preparedness efforts to reduce the
incidence of injuries and death in emergencies. “Planning is a process to manage risk” (Federal
Emergency Management Agency, 2010). Emergency plans are based on risk assessments. Risk
assessments are a foundational component of both comprehensive emergency management
programs and the national preparedness mission areas of prevention, protection, mitigation,
response, and recovery (U.S. Department of Homeland Security, 2013b). Incorporating socially
vulnerable populations within the risk management process ensures they are part of emergency
planning and resilience considerations. It is crucial that social vulnerability is included as part of
a jurisdiction’s risk assessment to direct mitigation efforts in cities. This ensures that mitigation
efforts go beyond building and environmental strategies and includes approaches to increase the
resilience of people and communities. It is also important to remember that it is the inter-
relationship between social characteristics, rather than an individual characteristic, that provides
a more accurate picture of vulnerability.
92
Based on this research, vulnerability varies across Boston neighborhoods. Despite these
challenges, we have significant opportunities for improvement to move away from inequities in
preparedness efforts caused by the false assumption of a “general population” and towards
planning that reflects the social conditions, characteristics, and needs of our communities. This
research presents the Social Determinants of Vulnerability Framework as a new tool to
strengthen national health security and increase social equity in mitigation, response, and
recovery. If Boston and other cities adopt this approach, we would have an equitable
development and execution of emergency plans that protects the rights of all people in our
communities. Rather than waiting for people to suffer after an emergency and incur the
economic and public health costs, using the Social Determinants of Vulnerability Framework is
an investment in building community resilience.
93
References
Assessing Department. (2014). Search FY2014 Real Estate Assessments and Taxes. Retrieved
6/15/2014, from City of Boston http://www.cityofboston.gov/assessing/
Boston Housing Authority. (2014). Development Information. Retrieved 6/15/2014, 2014, from
http://www.bostonhousing.org/housing_dev.html
Chandra, A., Acosta, J., Stern, S., Uscher-Pines, L., Williams, M. V., Yeung, D., . . . Meredith,
L. S. (2011). Building Community Relience to Disasters: A Way Forward to Enhance
National Health Security. Santa Monica: RAND Corporation.
City of Boston. (2014). Chinatown/Leather District. Retrieved 6/15/14, 2014, from
http://www.cityofboston.gov/neighborhoods/chinatownleatherdistrict.asp
Cutter, S. L., & Emrich, C. T. (2006). Moral Hazard, Social Catastrophe: The Changing Face of
Vulnerability along the Hurricane Coasts. Annals of the American Academy of Political
and Social Science, 604(1), 102-112.
Durant, T. J. (2011). The Utility of Vulnerability and Social Capital Theories in Studying the
Impact of Hurricane Katrina on the Elderly. Journal of Family Issues, 32(10), 1285-1302.
doi: 10.1177/0192513X11412491
Dwyer, A., Zoppou, C., Nielsen, O., Day, S., & Roberts, S. (2004). Quantifying Social
Vulnerability: A methodology for identifying those at risk to natural hazards. Geoscience
Australia Record, 2004/14.
Federal Emergency Management Agency. (2010). Comprehensive Preparedness Guide 101:
Developing and Maintaining Emergency Operations Plans, Version 2.0. Washington,
D.C.
94
Flanagan, B. E., Gregory, E. W., Hallisey, E. J., Heitgerd, J. L., & Lewis, B. (2011). A Social
Vulnerability Index for Disaster Management. Journal of Homeland Security and
Emergency Management: Vol. 8: Iss. 1, Article 3, 1-21.
Fothergill, A., Maestas, E. G. M., & Darlington, J. D. (1999). Race, Ethnicity and Disasters in
the United States: A Review of the Literature. Disasters, 23(2), 156-173. doi:
10.1111/1467-7717.00111
Fothergill, A., & Peek, L. A. (2004). Poverty and disasters in the United States: A review of
recent sociological findings. Natural Hazards, 32(1), 89-110.
Gustafsson, P. E., San Sebastian, M., Janlert, U., Theorell, T., Westerlund, H., & Hammarström,
A. (2014). Life-Course Accumulation of Neighborhood Disadvantage and Allostatic
Load: Empirical Integration of Three Social Determinants of Health Frameworks.
American Journal of Public Health, 104(5), 904-910.
Hawkins, R. L., & Maurer, K. (2009). Bonding, Bridging and Linking: How Social Capital
Operated in New Orleans following Hurricane Katrina. British Journal of Social Work,
40(6), 1777-1793. doi: 10.1093/bjsw/bcp087
Kailes, J. I., & Enders, A. (2007). Moving Beyond "Special Needs". Journal of Disability Policy
Studies, 17(4), 230-237.
MassHousing. (2014). Housing List. Boston, MA.
Mileti, D. S. (1999). Disasters by design: A reassessment of natural hazards in the United States.
Washington, D.C.: National Academies Press.
Milligan & Company, L., ,. (2007). Transportation Equity in Emergencies: A Review of the
Practices of State Departments of Transportation, Metropolitan Planning Organizations,
and Transit Agencies in 20 Metropolitan Areas (Vol. 2014).
95
Morrow, B. H. (1999). Identifying and Mapping Community Vulnerability. Disasters, 23(1), 1-
18.
Pantell, M., Rehkopf, D., Jutte, D., Syme, S. L., Balmes, J., & Adler, N. (2013). Social isolation:
a predictor of mortality comparable to traditional clinical risk factors. American Journal
of Public Health, 103(11), 2056-2062. doi: 10.2105/AJPH.2013.301261
Paulison, R. D. (2005). Civil Rights Program. Washington, D.C.: U.S. Department of Homeland
Security.
Rosso, P. (2012). Northampton Street project gets the OK from city, Boston Globe. Retrieved
from
http://www.boston.com/yourtown/news/roxbury/2012/09/northampton_street_project_get
.html
U.S. Census Bureau. (2012). 2012 ACS 5-Year Estimates. Retrieved from:
http://factfinder2.census.gov/faces/nav/jsf/pages/download_center.xhtml
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_12_5Y
R_DP04&prodType=table
U.S. Centers for Disease Control and Prevention. (2010). Public Health Workbook: To Define,
Locate, and Reach Special, Vulnerable, and At-Risk Populations in an Emergency.
Washington, D.C.
U.S. Department of Health and Human Services. (2009). National Health Security Strategy of
the United States of America. Washington, D.C.
U.S. Department of Homeland Security. (2013a). National Mitigation Framework. Washington,
D.C.
96
U.S. Department of Homeland Security. (2013b). NIPP 2013: Partnering for Critical
Infrastructure Security and Resilience. Washington, D.C.
U.S. Department of Homeland Security. (2014). 2014 Quadrennial Homeland Security Review.
Washington, D.C.: U.S. Department of Homeland Security.
97
Appendix A: Social Determinants of Vulnerability Frameworks
Social Determinants of Vulnerability Framework
98
Boston Social Determinants of Vulnerability Framework
99
Expanded Boston Social Determinants of Vulnerability Framework
100
100
Appendix B: Hot Spot Maps for each Social Determinant of Vulnerability
101
101
102
102
103
103
104
104
105
105
106
106
107
107
108
108
109
109
110
110
111
111
=
112
112
Appendix C: Neighborhood Correlation Analyses
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 0.933 0.489 0.846 0.685 0.851 0.684 .461 0.532 0.828 0.669 0.683
Sig. (2-tailed) .000 .039 .000 .002 .000 .002 .054 .023 .000 .002 .002
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.933 1 0.504 0.821 0.66 0.822 0.698 0.54 0.631 0.731 0.737 0.594
Sig. (2-tailed) .000 .033 .000 .003 .000 .001 .021 .005 .001 .000 .009
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.489 0.504 1 0.502 .339 .447 0.751 .412 .139 .223 .334 -.015
Sig. (2-tailed) .039 .033 .034 .169 .063 .000 .089 .583 .374 .176 .953
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.846 0.821 0.502 1 .324 0.592 .432 .041 .289 .456 .387 .321
Sig. (2-tailed) .000 .000 .034 .190 .010 .074 .873 .244 .057 .113 .194
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.685 0.66 .339 .324 1 0.954 0.708 0.724 0.548 0.867 0.683 0.835
Sig. (2-tailed) .002 .003 .169 .190 .000 .001 .001 .018 .000 .002 .000
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.851 0.822 .447 0.592 0.954 1 0.739 0.63 0.559 0.883 0.704 0.813
Sig. (2-tailed) .000 .000 .063 .010 .000 .000 .005 .016 .000 .001 .000
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.684 0.698 0.751 .432 0.708 0.739 1 0.756 .406 0.613 0.607 .413
Sig. (2-tailed) .002 .001 .000 .074 .001 .000 .000 .094 .007 .008 .088
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
.461 0.54 .412 .041 0.724 0.63 0.756 1 0.716 0.645 0.807 0.537
Sig. (2-tailed) .054 .021 .089 .873 .001 .005 .000 .001 .004 .000 .022
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.532 0.631 .139 .289 0.548 0.559 .406 0.716 1 0.665 0.959 0.59
Sig. (2-tailed) .023 .005 .583 .244 .018 .016 .094 .001 .003 .000 .010
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.828 0.731 .223 .456 0.867 0.883 0.613 0.645 0.665 1 0.765 0.933
Sig. (2-tailed) .000 .001 .374 .057 .000 .000 .007 .004 .003 .000 .000
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.669 0.737 .334 .387 0.683 0.704 0.607 0.807 0.959 0.765 1 0.621
Sig. (2-tailed) .002 .000 .176 .113 .002 .001 .008 .000 .000 .000 .006
N 18 18 18 18 18 18 18 18 18 18 18 18
Pearson Correlation
0.683 0.594 -.015 .321 0.835 0.813 .413 0.537 0.59 0.933 0.621 1
Sig. (2-tailed) .002 .009 .953 .194 .000 .000 .088 .022 .010 .000 .006
N 18 18 18 18 18 18 18 18 18 18 18 18
MedIllnes
NoVehicle
Low_to_No
lep
LessThanHS
poc
Women
occ_renter
Allston/Brighton Correlations
SocIsol
TotDis
TotChild
OlderAdult
113
113
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .747 .791 .808 .654 .940 .405 .871 .975 .942 .979 .899
Sig. (2-tailed) .013 .006 .005 .040 .000 .245 .001 .000 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
0.747 1.000 .494 .824 .493 .845 .346 .776 .635 .768 .660 .642
Sig. (2-tailed) .013 .147 .003 .148 .002 .327 .008 .049 .010 .038 .045
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.791 .494 1 .606 .348 .612 .025 .489 .813 .635 .824 .691
Sig. (2-tailed) .006 .147 .063 .325 .060 .946 .152 .004 .049 .003 .027
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.808 .824 .606 1 .204 .767 .046 .727 .699 .671 .748 .520
Sig. (2-tailed) .005 .003 .063 .572 .010 .901 .017 .025 .034 .013 .123
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.654 .493 .348 .204 1 .785 .722 .742 .702 .852 .653 .878
Sig. (2-tailed) .040 .148 .325 .572 .007 .018 .014 .024 .002 .041 .001
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.940 .845 .612 .767 .785 1 .502 .947 .903 .983 .902 .905
Sig. (2-tailed) .000 .002 .060 .010 .007 .139 .000 .000 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.405 .346 .025 .046 .722 .502 1 .542 .443 .584 .380 .565
Sig. (2-tailed) .245 .327 .946 .901 .018 .139 .106 .200 .076 .279 .089
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.871 .776 .489 .727 .742 .947 .542 1 .856 .938 .862 .782
Sig. (2-tailed) .001 .008 .152 .017 .014 .000 .106 .002 .000 .001 .008
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.975 .635 .813 .699 .702 .903 .443 .856 1 .933 .994 .906
Sig. (2-tailed) .000 .049 .004 .025 .024 .000 .200 .002 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.942 .768 .635 .671 .852 .983 .584 .938 .933 1 .919 .948
Sig. (2-tailed) .000 .010 .049 .034 .002 .000 .076 .000 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.979 .660 .824 .748 .653 .902 .380 .862 .994 .919 1 .874
Sig. (2-tailed) .000 .038 .003 .013 .041 .000 .279 .001 .000 .000 .001
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.899 .642 .691 .520 .878 .905 .565 .782 .906 .948 .874 1
Sig. (2-tailed) .000 .045 .027 .123 .001 .000 .089 .008 .000 .000 .001
N 10 10 10 10 10 10 10 10 10 10 10 10
LessThanHS
poc2
Women
occ_renter
MedIllnes
NoVehicle
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No
lep
Back Bay/Beacon Hill Correlations
114
114
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .746 .627 .442 .805 .854 .875 .811 .873 .969 .860 .956
Sig. (2-tailed) .089 .182 .380 .053 .030 .022 .050 .023 .001 .028 .003
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.746 1 .491 .695 .612 .711 .833 .625 .553 .763 .536 .662
Sig. (2-tailed) .089 .322 .125 .197 .113 .040 .185 .255 .078 .273 .152
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.627 .491 1 -.191 .632 .577 .421 .622 .548 .548 .536 .572
Sig. (2-tailed) .182 .322 .717 .178 .231 .406 .188 .260 .260 .273 .236
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.442 .695 -.191 1 .114 .284 .539 .123 .442 .491 .437 .325
Sig. (2-tailed) .380 .125 .717 .830 .585 .270 .817 .380 .322 .386 .530
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.805 .612 .632 .114 1 .985 .881 .997 .578 .872 .567 .923
Sig. (2-tailed) .053 .197 .178 .830 .000 .020 .000 .229 .023 .241 .009
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.854 .711 .577 .284 .985 1 .944 .983 .635 .927 .623 .947
Sig. (2-tailed) .030 .113 .231 .585 .000 .005 .000 .175 .008 .186 .004
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.875 .833 .421 .539 .881 .944 1 .896 .616 .933 .601 .922
Sig. (2-tailed) .022 .040 .406 .270 .020 .005 .016 .192 .007 .207 .009
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.811 .625 .622 .123 .997 .983 .896 1 .552 .868 .539 .927
Sig. (2-tailed) .050 .185 .188 .817 .000 .000 .016 .256 .025 .270 .008
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.873 .553 .548 .442 .578 .635 .616 .552 1 .852 1.000 .778
Sig. (2-tailed) .023 .255 .260 .380 .229 .175 .192 .256 .031 .000 .068
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.969 .763 .548 .491 .872 .927 .933 .868 .852 1 .842 .973
Sig. (2-tailed) .001 .078 .260 .322 .023 .008 .007 .025 .031 .035 .001
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.860 .536 .536 .437 .567 .623 .601 .539 1.000 .842 1 .766
Sig. (2-tailed) .028 .273 .273 .386 .241 .186 .207 .270 .000 .035 .076
N 6 6 6 6 6 6 6 6 6 6 6 6
Pearson Correlation
.956 .662 .572 .325 .923 .947 .922 .927 .778 .973 .766 1
Sig. (2-tailed) .003 .152 .236 .530 .009 .004 .009 .008 .068 .001 .076
N 6 6 6 6 6 6 6 6 6 6 6 6
Women
occ_renter
MedIllnes
NoVehicle
TotChild
OlderAdult
Low_to_No
lep
LessThanHS
poc2
Charlestwon Correlations
SocIsol
TotDis
115
115
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .914 .832 .952 .656 .837 .671 .687 .868 .931 .966 .873
Sig. (2-tailed) .001 .005 .000 .055 .005 .048 .041 .002 .000 .000 .002
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.914 1 .765 .953 .769 .915 .858 .706 .870 .799 .886 .764
Sig. (2-tailed) .001 .016 .000 .015 .001 .003 .034 .002 .010 .001 .017
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.832 .765 1 .711 .788 .830 .755 .877 .934 .833 .926 .717
Sig. (2-tailed) .005 .016 .032 .012 .006 .019 .002 .000 .005 .000 .030
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.952 .953 .711 1 .652 .853 .703 .613 .831 .838 .893 .818
Sig. (2-tailed) .000 .000 .032 .057 .003 .035 .080 .006 .005 .001 .007
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.656 .769 .788 .652 1 .952 .911 .918 .897 .673 .721 .606
Sig. (2-tailed) .055 .015 .012 .057 .000 .001 .000 .001 .047 .028 .084
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.837 .915 .830 .853 .952 1 .911 .880 .953 .802 .858 .748
Sig. (2-tailed) .005 .001 .006 .003 .000 .001 .002 .000 .009 .003 .020
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.671 .858 .755 .703 .911 .911 1 .846 .827 .556 .713 .485
Sig. (2-tailed) .048 .003 .019 .035 .001 .001 .004 .006 .120 .031 .186
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.687 .706 .877 .613 .918 .880 .846 1 .866 .675 .749 .526
Sig. (2-tailed) .041 .034 .002 .080 .000 .002 .004 .003 .046 .020 .146
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.868 .870 .934 .831 .897 .953 .827 .866 1 .882 .938 .827
Sig. (2-tailed) .002 .002 .000 .006 .001 .000 .006 .003 .002 .000 .006
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.931 .799 .833 .838 .673 .802 .556 .675 .882 1 .940 .963
Sig. (2-tailed) .000 .010 .005 .005 .047 .009 .120 .046 .002 .000 .000
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.966 .886 .926 .893 .721 .858 .713 .749 .938 .940 1 .887
Sig. (2-tailed) .000 .001 .000 .001 .028 .003 .031 .020 .000 .000 .001
N 9 9 9 9 9 9 9 9 9 9 9 9
Pearson Correlation
.873 .764 .717 .818 .606 .748 .485 .526 .827 .963 .887 1
Sig. (2-tailed) .002 .017 .030 .007 .084 .020 .186 .146 .006 .000 .001
N 9 9 9 9 9 9 9 9 9 9 9 9
LessThanHS
poc2
Women
occ_renter
MedIllnes
NoVehicle
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No
lep
Downtown Correlations
116
116
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .925 .974 .843 .918 .954 .776 .887 .982 .982 .974 .929
Sig. (2-tailed) .000 .000 .000 .000 .000 .001 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.925 1 .890 .882 .851 .911 .664 .765 .863 .895 .844 .844
Sig. (2-tailed) .000 .000 .000 .000 .000 .010 .001 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.974 .890 1 .777 .955 .966 .831 .920 .976 .957 .968 .923
Sig. (2-tailed) .000 .000 .001 .000 .000 .000 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.843 .882 .777 1 .703 .825 .421 .547 .792 .786 .762 .756
Sig. (2-tailed) .000 .000 .001 .005 .000 .134 .043 .001 .001 .002 .002
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.918 .851 .955 .703 1 .982 .853 .919 .900 .923 .897 .964
Sig. (2-tailed) .000 .000 .000 .005 .000 .000 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.954 .911 .966 .825 .982 1 .790 .875 .926 .943 .916 .967
Sig. (2-tailed) .000 .000 .000 .000 .000 .001 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.776 .664 .831 .421 .853 .790 1 .956 .791 .838 .819 .827
Sig. (2-tailed) .001 .010 .000 .134 .000 .001 .000 .001 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.887 .765 .920 .547 .919 .875 .956 1 .894 .928 .911 .902
Sig. (2-tailed) .000 .001 .000 .043 .000 .000 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.982 .863 .976 .792 .900 .926 .791 .894 1 .964 .997 .897
Sig. (2-tailed) .000 .000 .000 .001 .000 .000 .001 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.982 .895 .957 .786 .923 .943 .838 .928 .964 1 .965 .957
Sig. (2-tailed) .000 .000 .000 .001 .000 .000 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.974 .844 .968 .762 .897 .916 .819 .911 .997 .965 1 .899
Sig. (2-tailed) .000 .000 .000 .002 .000 .000 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.929 .844 .923 .756 .964 .967 .827 .902 .897 .957 .899 1
Sig. (2-tailed) .000 .000 .000 .002 .000 .000 .000 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
MedIllnes
NoVehicle
Low_to_No
lep
LessThanHS
poc2
Women
occ_renter
East Boston Correlations
SocIsol
TotDis
TotChild
OlderAdult
117
117
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .760 .444 .738 .701 .808 .595 .262 -.027 .922 .305 .940
Sig. (2-tailed) .011 .199 .015 .024 .005 .069 .465 .941 .000 .391 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.760 1 .335 .768 .606 .730 .728 .418 .222 .524 .498 .555
Sig. (2-tailed) .011 .344 .010 .063 .016 .017 .229 .538 .120 .143 .096
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.444 .335 1 .053 .298 .279 .055 .526 .607 .452 .655 .446
Sig. (2-tailed) .199 .344 .884 .403 .435 .879 .118 .063 .190 .040 .197
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.738 .768 .053 1 .326 .538 .782 .046 -.221 .453 .053 .523
Sig. (2-tailed) .015 .010 .884 .357 .108 .008 .899 .540 .189 .884 .121
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.701 .606 .298 .326 1 .972 .332 .357 .176 .774 .474 .764
Sig. (2-tailed) .024 .063 .403 .357 .000 .349 .311 .627 .009 .166 .010
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.808 .730 .279 .538 .972 1 .489 .330 .102 .802 .436 .810
Sig. (2-tailed) .005 .016 .435 .108 .000 .151 .352 .779 .005 .208 .004
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.595 .728 .055 .782 .332 .489 1 -.005 -.303 .352 -.072 .425
Sig. (2-tailed) .069 .017 .879 .008 .349 .151 .988 .395 .318 .842 .221
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.262 .418 .526 .046 .357 .330 -.005 1 .868 .256 .897 .278
Sig. (2-tailed) .465 .229 .118 .899 .311 .352 .988 .001 .475 .000 .437
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
-.027 .222 .607 -.221 .176 .102 -.303 .868 1 -.017 .926 -.021
Sig. (2-tailed) .941 .538 .063 .540 .627 .779 .395 .001 .963 .000 .953
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.922 .524 .452 .453 .774 .802 .352 .256 -.017 1 .280 .992
Sig. (2-tailed) .000 .120 .190 .189 .009 .005 .318 .475 .963 .433 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.305 .498 .655 .053 .474 .436 -.072 .897 .926 .280 1 .282
Sig. (2-tailed) .391 .143 .040 .884 .166 .208 .842 .000 .000 .433 .429
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.940 .555 .446 .523 .764 .810 .425 .278 -.021 .992 .282 1
Sig. (2-tailed) .000 .096 .197 .121 .010 .004 .221 .437 .953 .000 .429
N 10 10 10 10 10 10 10 10 10 10 10 10
Women
occ_renter
MedIllnes
NoVehicle
TotChild
OlderAdult
Low_to_No
lep
LessThanHS
poc2
Fenway/Kenmore Correlations
SocIsol
TotDis
118
118
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .952 .864 .988 .971 .986 .886 .971 .987 .943 .988 .885
Sig. (2-tailed) .000 .006 .000 .000 .000 .003 .000 .000 .000 .000 .004
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.952 1 .790 .968 .912 .942 .740 .876 .932 .907 .942 .867
Sig. (2-tailed) .000 .020 .000 .002 .000 .036 .004 .001 .002 .000 .005
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.864 .790 1 .853 .868 .869 .881 .873 .932 .668 .927 .566
Sig. (2-tailed) .006 .020 .007 .005 .005 .004 .005 .001 .070 .001 .143
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.988 .968 .853 1 .967 .988 .866 .957 .973 .917 .977 .895
Sig. (2-tailed) .000 .000 .007 .000 .000 .005 .000 .000 .001 .000 .003
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.971 .912 .868 .967 1 .995 .934 .989 .972 .875 .968 .819
Sig. (2-tailed) .000 .002 .005 .000 .000 .001 .000 .000 .004 .000 .013
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.986 .942 .869 .988 .995 1 .914 .984 .980 .899 .979 .856
Sig. (2-tailed) .000 .000 .005 .000 .000 .001 .000 .000 .002 .000 .007
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.886 .740 .881 .866 .934 .914 1 .966 .909 .760 .899 .698
Sig. (2-tailed) .003 .036 .004 .005 .001 .001 .000 .002 .029 .002 .054
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.971 .876 .873 .957 .989 .984 .966 1 .969 .884 .964 .829
Sig. (2-tailed) .000 .004 .005 .000 .000 .000 .000 .000 .004 .000 .011
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.987 .932 .932 .973 .972 .980 .909 .969 1 .883 .999 .803
Sig. (2-tailed) .000 .001 .001 .000 .000 .000 .002 .000 .004 .000 .016
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.943 .907 .668 .917 .875 .899 .760 .884 .883 1 .888 .956
Sig. (2-tailed) .000 .002 .070 .001 .004 .002 .029 .004 .004 .003 .000
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.988 .942 .927 .977 .968 .979 .899 .964 .999 .888 1 .810
Sig. (2-tailed) .000 .000 .001 .000 .000 .000 .002 .000 .000 .003 .015
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.885 .867 .566 .895 .819 .856 .698 .829 .803 .956 .810 1
Sig. (2-tailed) .004 .005 .143 .003 .013 .007 .054 .011 .016 .000 .015
N 8 8 8 8 8 8 8 8 8 8 8 8
LessThanHS
poc2
Women
occ_renter
MedIllnes
NoVehicle
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No
lep
Hyde Park Correlations
119
119
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .918 .825 .680 .886 .905 .769 .863 .903 .892 .893 .866
Sig. (2-tailed) .000 .000 .005 .000 .000 .001 .000 .000 .000 .000 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.918 1 .766 .637 .898 .907 .832 .886 .741 .808 .727 .818
Sig. (2-tailed) .000 .001 .011 .000 .000 .000 .000 .002 .000 .002 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.825 .766 1 .454 .706 .704 .699 .812 .622 .543 .613 .529
Sig. (2-tailed) .000 .001 .089 .003 .003 .004 .000 .013 .036 .015 .043
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.680 .637 .454 1 .596 .712 .529 .551 .755 .632 .782 .683
Sig. (2-tailed) .005 .011 .089 .019 .003 .042 .033 .001 .011 .001 .005
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.886 .898 .706 .596 1 .988 .838 .943 .728 .898 .728 .930
Sig. (2-tailed) .000 .000 .003 .019 .000 .000 .000 .002 .000 .002 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.905 .907 .704 .712 .988 1 .834 .930 .781 .906 .786 .944
Sig. (2-tailed) .000 .000 .003 .003 .000 .000 .000 .001 .000 .001 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.769 .832 .699 .529 .838 .834 1 .934 .580 .680 .589 .705
Sig. (2-tailed) .001 .000 .004 .042 .000 .000 .000 .024 .005 .021 .003
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.863 .886 .812 .551 .943 .930 .934 1 .680 .774 .682 .794
Sig. (2-tailed) .000 .000 .000 .033 .000 .000 .000 .005 .001 .005 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.903 .741 .622 .755 .728 .781 .580 .680 1 .874 .997 .808
Sig. (2-tailed) .000 .002 .013 .001 .002 .001 .024 .005 .000 .000 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.892 .808 .543 .632 .898 .906 .680 .774 .874 1 .867 .975
Sig. (2-tailed) .000 .000 .036 .011 .000 .000 .005 .001 .000 .000 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.893 .727 .613 .782 .728 .786 .589 .682 .997 .867 1 .808
Sig. (2-tailed) .000 .002 .015 .001 .002 .001 .021 .005 .000 .000 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
Pearson Correlation
.866 .818 .529 .683 .930 .944 .705 .794 .808 .975 .808 1
Sig. (2-tailed) .000 .000 .043 .005 .000 .000 .003 .000 .000 .000 .000
N 15 15 15 15 15 15 15 15 15 15 15 15
MedIllnes
NoVehicle
Low_to_No
lep
LessThanHS
poc2
Women
occ_renter
Jamaica Plain Correlations
SocIsol
TotDis
TotChild
OlderAdult
120
120
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .918 .792 .768 .781 .868 .774 .977 .947 .986 .923 .968
Sig. (2-tailed) .001 .019 .026 .022 .005 .024 .000 .000 .000 .001 .000
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.918 1 .712 .750 .780 .863 .898 .966 .897 .908 .899 .894
Sig. (2-tailed) .001 .048 .032 .022 .006 .002 .000 .003 .002 .002 .003
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.792 .712 1 .291 .948 .909 .767 .761 .628 .831 .578 .782
Sig. (2-tailed) .019 .048 .485 .000 .002 .027 .028 .095 .011 .134 .022
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.768 .750 .291 1 .396 .578 .514 .749 .867 .729 .891 .654
Sig. (2-tailed) .026 .032 .485 .331 .134 .192 .032 .005 .040 .003 .079
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.781 .780 .948 .396 1 .978 .849 .774 .686 .833 .656 .740
Sig. (2-tailed) .022 .022 .000 .331 .000 .008 .024 .060 .010 .077 .036
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.868 .863 .909 .578 .978 1 .871 .857 .806 .905 .784 .805
Sig. (2-tailed) .005 .006 .002 .134 .000 .005 .007 .016 .002 .021 .016
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.774 .898 .767 .514 .849 .871 1 .855 .759 .767 .772 .796
Sig. (2-tailed) .024 .002 .027 .192 .008 .005 .007 .029 .026 .025 .018
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.977 .966 .761 .749 .774 .857 .855 1 .934 .962 .925 .974
Sig. (2-tailed) .000 .000 .028 .032 .024 .007 .007 .001 .000 .001 .000
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.947 .897 .628 .867 .686 .806 .759 .934 1 .910 .994 .908
Sig. (2-tailed) .000 .003 .095 .005 .060 .016 .029 .001 .002 .000 .002
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.986 .908 .831 .729 .833 .905 .767 .962 .910 1 .883 .943
Sig. (2-tailed) .000 .002 .011 .040 .010 .002 .026 .000 .002 .004 .000
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.923 .899 .578 .891 .656 .784 .772 .925 .994 .883 1 .885
Sig. (2-tailed) .001 .002 .134 .003 .077 .021 .025 .001 .000 .004 .003
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.968 .894 .782 .654 .740 .805 .796 .974 .908 .943 .885 1
Sig. (2-tailed) .000 .003 .022 .079 .036 .016 .018 .000 .002 .000 .003
N 8 8 8 8 8 8 8 8 8 8 8 8
Women
occ_renter
MedIllnes
NoVehicle
TotChild
OlderAdult
Low_to_No
lep
LessThanHS
poc2
Mattapan Correlations
SocIsol
TotDis
121
121
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .960 .825 .893 .233 .425 .770 .534 .746 .394 .836 .689
Sig. (2-tailed) .000 .012 .003 .579 .294 .025 .173 .034 .333 .010 .059
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.960 1 .881 .837 .421 .593 .738 .568 .778 .506 .810 .798
Sig. (2-tailed) .000 .004 .010 .299 .121 .037 .141 .023 .201 .015 .018
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.825 .881 1 .759 .423 .578 .784 .756 .717 .400 .717 .579
Sig. (2-tailed) .012 .004 .029 .296 .134 .021 .030 .045 .326 .046 .132
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.893 .837 .759 1 .054 .277 .833 .327 .647 .212 .778 .448
Sig. (2-tailed) .003 .010 .029 .899 .506 .010 .429 .083 .614 .023 .265
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.233 .421 .423 .054 1 .974 -.140 .331 .627 .863 .419 .787
Sig. (2-tailed) .579 .299 .296 .899 .000 .741 .423 .096 .006 .302 .020
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.425 .593 .578 .277 .974 1 .053 .392 .749 .878 .578 .858
Sig. (2-tailed) .294 .121 .134 .506 .000 .901 .337 .032 .004 .133 .006
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.770 .738 .784 .833 -.140 .053 1 .603 .295 -.162 .426 .208
Sig. (2-tailed) .025 .037 .021 .010 .741 .901 .114 .478 .701 .292 .621
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.534 .568 .756 .327 .331 .392 .603 1 .306 .092 .264 .363
Sig. (2-tailed) .173 .141 .030 .429 .423 .337 .114 .462 .828 .528 .377
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.746 .778 .717 .647 .627 .749 .295 .306 1 .829 .966 .791
Sig. (2-tailed) .034 .023 .045 .083 .096 .032 .478 .462 .011 .000 .019
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.394 .506 .400 .212 .863 .878 -.162 .092 .829 1 .692 .837
Sig. (2-tailed) .333 .201 .326 .614 .006 .004 .701 .828 .011 .057 .010
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.836 .810 .717 .778 .419 .578 .426 .264 .966 .692 1 .704
Sig. (2-tailed) .010 .015 .046 .023 .302 .133 .292 .528 .000 .057 .051
N 8 8 8 8 8 8 8 8 8 8 8 8
Pearson Correlation
.689 .798 .579 .448 .787 .858 .208 .363 .791 .837 .704 1
Sig. (2-tailed) .059 .018 .132 .265 .020 .006 .621 .377 .019 .010 .051
N 8 8 8 8 8 8 8 8 8 8 8 8
LessThanHS
poc2
Women
occ_renter
MedIllnes
NoVehicle
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No
lep
North Dorchester Correlations
122
122
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .845 .913 .664 .813 .939 .350 .741 .954 .935 .931 .859
Sig. (2-tailed) .002 .000 .036 .004 .000 .322 .014 .000 .000 .000 .001
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.845 1 .935 .637 .723 .858 .475 .632 .847 .675 .836 .796
Sig. (2-tailed) .002 .000 .048 .018 .001 .165 .050 .002 .032 .003 .006
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.913 .935 1 .501 .906 .928 .457 .797 .862 .843 .826 .899
Sig. (2-tailed) .000 .000 .140 .000 .000 .184 .006 .001 .002 .003 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.664 .637 .501 1 .257 .690 .172 .032 .792 .503 .834 .278
Sig. (2-tailed) .036 .048 .140 .474 .027 .636 .930 .006 .139 .003 .437
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.813 .723 .906 .257 1 .877 .330 .851 .698 .881 .633 .887
Sig. (2-tailed) .004 .018 .000 .474 .001 .352 .002 .025 .001 .049 .001
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.939 .858 .928 .690 .877 1 .332 .654 .917 .910 .889 .803
Sig. (2-tailed) .000 .001 .000 .027 .001 .348 .040 .000 .000 .001 .005
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.350 .475 .457 .172 .330 .332 1 .345 .293 .264 .281 .288
Sig. (2-tailed) .322 .165 .184 .636 .352 .348 .328 .411 .460 .432 .420
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.741 .632 .797 .032 .851 .654 .345 1 .586 .737 .521 .908
Sig. (2-tailed) .014 .050 .006 .930 .002 .040 .328 .075 .015 .123 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.954 .847 .862 .792 .698 .917 .293 .586 1 .842 .995 .728
Sig. (2-tailed) .000 .002 .001 .006 .025 .000 .411 .075 .002 .000 .017
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.935 .675 .843 .503 .881 .910 .264 .737 .842 1 .796 .848
Sig. (2-tailed) .000 .032 .002 .139 .001 .000 .460 .015 .002 .006 .002
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.931 .836 .826 .834 .633 .889 .281 .521 .995 .796 1 .669
Sig. (2-tailed) .000 .003 .003 .003 .049 .001 .432 .123 .000 .006 .034
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.859 .796 .899 .278 .887 .803 .288 .908 .728 .848 .669 1
Sig. (2-tailed) .001 .006 .000 .437 .001 .005 .420 .000 .017 .002 .034
N 10 10 10 10 10 10 10 10 10 10 10 10
MedIllnes
NoVehicle
Low_to_No
lep
LessThanHS
poc2
Women
occ_renter
Roslindale Correlations
SocIsol
TotDis
TotChild
OlderAdult
123
123
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .845 .913 .664 .813 .939 .350 .741 .954 .935 .931 .859
Sig. (2-tailed) .002 .000 .036 .004 .000 .322 .014 .000 .000 .000 .001
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.845 1 .935 .637 .723 .858 .475 .632 .847 .675 .836 .796
Sig. (2-tailed) .002 .000 .048 .018 .001 .165 .050 .002 .032 .003 .006
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.913 .935 1 .501 .906 .928 .457 .797 .862 .843 .826 .899
Sig. (2-tailed) .000 .000 .140 .000 .000 .184 .006 .001 .002 .003 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.664 .637 .501 1 .257 .690 .172 .032 .792 .503 .834 .278
Sig. (2-tailed) .036 .048 .140 .474 .027 .636 .930 .006 .139 .003 .437
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.813 .723 .906 .257 1 .877 .330 .851 .698 .881 .633 .887
Sig. (2-tailed) .004 .018 .000 .474 .001 .352 .002 .025 .001 .049 .001
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.939 .858 .928 .690 .877 1 .332 .654 .917 .910 .889 .803
Sig. (2-tailed) .000 .001 .000 .027 .001 .348 .040 .000 .000 .001 .005
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.350 .475 .457 .172 .330 .332 1 .345 .293 .264 .281 .288
Sig. (2-tailed) .322 .165 .184 .636 .352 .348 .328 .411 .460 .432 .420
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.741 .632 .797 .032 .851 .654 .345 1 .586 .737 .521 .908
Sig. (2-tailed) .014 .050 .006 .930 .002 .040 .328 .075 .015 .123 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.954 .847 .862 .792 .698 .917 .293 .586 1 .842 .995 .728
Sig. (2-tailed) .000 .002 .001 .006 .025 .000 .411 .075 .002 .000 .017
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.935 .675 .843 .503 .881 .910 .264 .737 .842 1 .796 .848
Sig. (2-tailed) .000 .032 .002 .139 .001 .000 .460 .015 .002 .006 .002
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.931 .836 .826 .834 .633 .889 .281 .521 .995 .796 1 .669
Sig. (2-tailed) .000 .003 .003 .003 .049 .001 .432 .123 .000 .006 .034
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.859 .796 .899 .278 .887 .803 .288 .908 .728 .848 .669 1
Sig. (2-tailed) .001 .006 .000 .437 .001 .005 .420 .000 .017 .002 .034
N 10 10 10 10 10 10 10 10 10 10 10 10
MedIllnes
NoVehicle
Low_to_No
lep
LessThanHS
poc2
Women
occ_renter
Roslindale Correlations
SocIsol
TotDis
TotChild
OlderAdult
124
124
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .931 .916 .853 .871 .901 .720 .934 .802 .973 .771 .891
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000 .000 .000 .000 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.931 1 .839 .883 .834 .876 .782 .858 .659 .893 .638 .827
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000 .002 .000 .002 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.916 .839 1 .708 .832 .841 .724 .936 .754 .854 .724 .751
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000 .000 .000 .000 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.853 .883 .708 1 .754 .829 .730 .742 .594 .828 .607 .809
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000 .006 .000 .005 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.871 .834 .832 .754 1 .992 .699 .787 .663 .878 .614 .898
Sig. (2-tailed) .000 .000 .000 .000 .000 .001 .000 .001 .000 .004 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.901 .876 .841 .829 .992 1 .732 .809 .676 .902 .636 .916
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000 .001 .000 .003 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.720 .782 .724 .730 .699 .732 1 .795 .580 .725 .679 .640
Sig. (2-tailed) .000 .000 .000 .000 .001 .000 .000 .007 .000 .001 .002
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.934 .858 .936 .742 .787 .809 .795 1 .823 .896 .847 .777
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000 .000 .000 .000 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.802 .659 .754 .594 .663 .676 .580 .823 1 .855 .964 .729
Sig. (2-tailed) .000 .002 .000 .006 .001 .001 .007 .000 .000 .000 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.973 .893 .854 .828 .878 .902 .725 .896 .855 1 .830 .920
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .000 .000 .000 .000 .000
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.771 .638 .724 .607 .614 .636 .679 .847 .964 .830 1 .694
Sig. (2-tailed) .000 .002 .000 .005 .004 .003 .001 .000 .000 .000 .001
N 20 20 20 20 20 20 20 20 20 20 20 20
Pearson Correlation
.891 .827 .751 .809 .898 .916 .640 .777 .729 .920 .694 1
Sig. (2-tailed) .000 .000 .000 .000 .000 .000 .002 .000 .000 .000 .001
N 20 20 20 20 20 20 20 20 20 20 20 20
Women
occ_renter
MedIllnes
NoVehicle
TotChild
OlderAdult
Low_to_No
lep
LessThanHS
poc2
Roxbury Correlations
SocIsol
TotDis
125
125
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .725 .649 .805 .653 .823 .735 .473 .849 .943 .817 .877
Sig. (2-tailed) .005 .016 .001 .015 .001 .004 .103 .000 .000 .001 .000
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.725 1 .845 .408 .851 .895 .789 .687 .402 .578 .361 .673
Sig. (2-tailed) .005 .000 .167 .000 .000 .001 .009 .173 .039 .225 .012
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.649 .845 1 .187 .961 .935 .896 .886 .243 .490 .185 .633
Sig. (2-tailed) .016 .000 .542 .000 .000 .000 .000 .423 .089 .544 .020
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.805 .408 .187 1 .155 .420 .331 -.064 .910 .771 .894 .544
Sig. (2-tailed) .001 .167 .542 .612 .153 .270 .835 .000 .002 .000 .055
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.653 .851 .961 .155 1 .962 .929 .947 .194 .482 .143 .689
Sig. (2-tailed) .015 .000 .000 .612 .000 .000 .000 .526 .095 .641 .009
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.823 .895 .935 .420 .962 1 .945 .852 .430 .657 .379 .783
Sig. (2-tailed) .001 .000 .000 .153 .000 .000 .000 .142 .015 .202 .002
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.735 .789 .896 .331 .929 .945 1 .876 .294 .561 .234 .705
Sig. (2-tailed) .004 .001 .000 .270 .000 .000 .000 .329 .046 .442 .007
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.473 .687 .886 -.064 .947 .852 .876 1 -.032 .317 -.084 .582
Sig. (2-tailed) .103 .009 .000 .835 .000 .000 .000 .918 .291 .784 .037
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.849 .402 .243 .910 .194 .430 .294 -.032 1 .909 .996 .689
Sig. (2-tailed) .000 .173 .423 .000 .526 .142 .329 .918 .000 .000 .009
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.943 .578 .490 .771 .482 .657 .561 .317 .909 1 .894 .902
Sig. (2-tailed) .000 .039 .089 .002 .095 .015 .046 .291 .000 .000 .000
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.817 .361 .185 .894 .143 .379 .234 -.084 .996 .894 1 .665
Sig. (2-tailed) .001 .225 .544 .000 .641 .202 .442 .784 .000 .000 .013
N 13 13 13 13 13 13 13 13 13 13 13 13
Pearson Correlation
.877 .673 .633 .544 .689 .783 .705 .582 .689 .902 .665 1
Sig. (2-tailed) .000 .012 .020 .055 .009 .002 .007 .037 .009 .000 .013
N 13 13 13 13 13 13 13 13 13 13 13 13
LessThanHS
poc2
Women
occ_renter
MedIllnes
NoVehicle
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No
lep
South Boston Correlations
126
126
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .859 .815 .704 .375 .551 .702 .623 .969 .979 .963 .871
Sig. (2-tailed) .000 .000 .005 .187 .041 .005 .017 .000 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.859 1 .875 .589 .660 .795 .792 .689 .818 .873 .788 .821
Sig. (2-tailed) .000 .000 .027 .010 .001 .001 .006 .000 .000 .001 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.815 .875 1 .498 .666 .775 .710 .671 .729 .821 .726 .831
Sig. (2-tailed) .000 .000 .070 .009 .001 .004 .009 .003 .000 .003 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.704 .589 .498 1 .000 .270 .236 .020 .739 .673 .774 .434
Sig. (2-tailed) .005 .027 .070 1.000 .351 .417 .947 .003 .008 .001 .121
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.375 .660 .666 .000 1 .963 .790 .682 .358 .431 .310 .405
Sig. (2-tailed) .187 .010 .009 1.000 .000 .001 .007 .209 .124 .281 .151
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.551 .795 .775 .270 .963 1 .824 .662 .544 .597 .508 .507
Sig. (2-tailed) .041 .001 .001 .351 .000 .000 .010 .044 .024 .064 .064
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.702 .792 .710 .236 .790 .824 1 .741 .657 .751 .616 .679
Sig. (2-tailed) .005 .001 .004 .417 .001 .000 .002 .011 .002 .019 .008
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.623 .689 .671 .020 .682 .662 .741 1 .561 .614 .502 .705
Sig. (2-tailed) .017 .006 .009 .947 .007 .010 .002 .037 .020 .067 .005
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.969 .818 .729 .739 .358 .544 .657 .561 1 .960 .995 .780
Sig. (2-tailed) .000 .000 .003 .003 .209 .044 .011 .037 .000 .000 .001
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.979 .873 .821 .673 .431 .597 .751 .614 .960 1 .951 .885
Sig. (2-tailed) .000 .000 .000 .008 .124 .024 .002 .020 .000 .000 .000
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.963 .788 .726 .774 .310 .508 .616 .502 .995 .951 1 .763
Sig. (2-tailed) .000 .001 .003 .001 .281 .064 .019 .067 .000 .000 .001
N 14 14 14 14 14 14 14 14 14 14 14 14
Pearson Correlation
.871 .821 .831 .434 .405 .507 .679 .705 .780 .885 .763 1
Sig. (2-tailed) .000 .000 .000 .121 .151 .064 .008 .005 .001 .000 .001
N 14 14 14 14 14 14 14 14 14 14 14 14
MedIllnes
NoVehicle
Low_to_No
lep
LessThanHS
poc2
Women
occ_renter
South Dorchester Correlations
SocIsol
TotDis
TotChild
OlderAdult
127
127
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .556 .730 .535 .667 .729 .473 .698 .878 .901 .833 .854
Sig. (2-tailed) .095 .017 .111 .035 .017 .168 .025 .001 .000 .003 .002
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.556 1 .651 .138 .870 .798 .704 .840 .289 .382 .249 .320
Sig. (2-tailed) .095 .041 .704 .001 .006 .023 .002 .417 .276 .488 .367
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.730 .651 1 .503 .697 .747 .712 .661 .641 .494 .612 .419
Sig. (2-tailed) .017 .041 .139 .025 .013 .021 .037 .046 .146 .060 .228
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.535 .138 .503 1 .335 .567 .602 .284 .644 .526 .609 .414
Sig. (2-tailed) .111 .704 .139 .344 .087 .065 .426 .045 .118 .062 .235
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.667 .870 .697 .335 1 .966 .729 .880 .419 .572 .364 .400
Sig. (2-tailed) .035 .001 .025 .344 .000 .017 .001 .229 .084 .302 .252
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.729 .798 .747 .567 .966 1 .802 .847 .542 .644 .485 .463
Sig. (2-tailed) .017 .006 .013 .087 .000 .005 .002 .105 .044 .156 .178
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.473 .704 .712 .602 .729 .802 1 .793 .340 .323 .261 .147
Sig. (2-tailed) .168 .023 .021 .065 .017 .005 .006 .337 .363 .466 .686
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.698 .840 .661 .284 .880 .847 .793 1 .386 .544 .285 .386
Sig. (2-tailed) .025 .002 .037 .426 .001 .002 .006 .271 .104 .425 .270
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.878 .289 .641 .644 .419 .542 .340 .386 1 .910 .986 .915
Sig. (2-tailed) .001 .417 .046 .045 .229 .105 .337 .271 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.901 .382 .494 .526 .572 .644 .323 .544 .910 1 .873 .953
Sig. (2-tailed) .000 .276 .146 .118 .084 .044 .363 .104 .000 .001 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.833 .249 .612 .609 .364 .485 .261 .285 .986 .873 1 .902
Sig. (2-tailed) .003 .488 .060 .062 .302 .156 .466 .425 .000 .001 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.854 .320 .419 .414 .400 .463 .147 .386 .915 .953 .902 1
Sig. (2-tailed) .002 .367 .228 .235 .252 .178 .686 .270 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Women
occ_renter
MedIllnes
NoVehicle
TotChild
OlderAdult
Low_to_No
lep
LessThanHS
poc2
South End Correlations
SocIsol
TotDis
128
128
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .556 .730 .535 .667 .729 .473 .698 .878 .901 .833 .854
Sig. (2-tailed) .095 .017 .111 .035 .017 .168 .025 .001 .000 .003 .002
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.556 1 .651 .138 .870 .798 .704 .840 .289 .382 .249 .320
Sig. (2-tailed) .095 .041 .704 .001 .006 .023 .002 .417 .276 .488 .367
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.730 .651 1 .503 .697 .747 .712 .661 .641 .494 .612 .419
Sig. (2-tailed) .017 .041 .139 .025 .013 .021 .037 .046 .146 .060 .228
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.535 .138 .503 1 .335 .567 .602 .284 .644 .526 .609 .414
Sig. (2-tailed) .111 .704 .139 .344 .087 .065 .426 .045 .118 .062 .235
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.667 .870 .697 .335 1 .966 .729 .880 .419 .572 .364 .400
Sig. (2-tailed) .035 .001 .025 .344 .000 .017 .001 .229 .084 .302 .252
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.729 .798 .747 .567 .966 1 .802 .847 .542 .644 .485 .463
Sig. (2-tailed) .017 .006 .013 .087 .000 .005 .002 .105 .044 .156 .178
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.473 .704 .712 .602 .729 .802 1 .793 .340 .323 .261 .147
Sig. (2-tailed) .168 .023 .021 .065 .017 .005 .006 .337 .363 .466 .686
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.698 .840 .661 .284 .880 .847 .793 1 .386 .544 .285 .386
Sig. (2-tailed) .025 .002 .037 .426 .001 .002 .006 .271 .104 .425 .270
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.878 .289 .641 .644 .419 .542 .340 .386 1 .910 .986 .915
Sig. (2-tailed) .001 .417 .046 .045 .229 .105 .337 .271 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.901 .382 .494 .526 .572 .644 .323 .544 .910 1 .873 .953
Sig. (2-tailed) .000 .276 .146 .118 .084 .044 .363 .104 .000 .001 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.833 .249 .612 .609 .364 .485 .261 .285 .986 .873 1 .902
Sig. (2-tailed) .003 .488 .060 .062 .302 .156 .466 .425 .000 .001 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Pearson Correlation
.854 .320 .419 .414 .400 .463 .147 .386 .915 .953 .902 1
Sig. (2-tailed) .002 .367 .228 .235 .252 .178 .686 .270 .000 .000 .000
N 10 10 10 10 10 10 10 10 10 10 10 10
Women
occ_renter
MedIllnes
NoVehicle
TotChild
OlderAdult
Low_to_No
lep
LessThanHS
poc2
South End Correlations
SocIsol
TotDis
129
129
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No lep
LessThanHS poc Women
occ_renter
MedIllnes
NoVehicle
Pearson Correlation
1 .516 .789 -.004 .910 .972 .745 .918 .597 .956 .504 .582
Sig. (2-tailed) .235 .035 .993 .004 .000 .055 .004 .157 .001 .249 .170
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.516 1 .318 .719 .136 .330 .078 .178 .773 .312 .764 .003
Sig. (2-tailed) .235 .487 .068 .772 .469 .868 .702 .041 .496 .046 .996
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.789 .318 1 -.019 .726 .771 .453 .773 .754 .784 .683 .166
Sig. (2-tailed) .035 .487 .968 .065 .042 .308 .042 .050 .037 .091 .722
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
-.004 .719 -.019 1 -.380 -.149 -.486 -.361 .573 -.265 .613 -.610
Sig. (2-tailed) .993 .068 .968 .400 .750 .268 .426 .179 .566 .143 .146
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.910 .136 .726 -.380 1 .971 .832 .994 .285 .969 .169 .746
Sig. (2-tailed) .004 .772 .065 .400 .000 .020 .000 .535 .000 .718 .054
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.972 .330 .771 -.149 .971 1 .764 .970 .453 .968 .338 .640
Sig. (2-tailed) .000 .469 .042 .750 .000 .045 .000 .308 .000 .458 .122
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.745 .078 .453 -.486 .832 .764 1 .807 .124 .855 .066 .803
Sig. (2-tailed) .055 .868 .308 .268 .020 .045 .028 .791 .014 .889 .030
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.918 .178 .773 -.361 .994 .970 .807 1 .349 .979 .229 .723
Sig. (2-tailed) .004 .702 .042 .426 .000 .000 .028 .443 .000 .622 .066
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.597 .773 .754 .573 .285 .453 .124 .349 1 .460 .986 -.209
Sig. (2-tailed) .157 .041 .050 .179 .535 .308 .791 .443 .299 .000 .652
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.956 .312 .784 -.265 .969 .968 .855 .979 .460 1 .355 .699
Sig. (2-tailed) .001 .496 .037 .566 .000 .000 .014 .000 .299 .435 .081
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.504 .764 .683 .613 .169 .338 .066 .229 .986 .355 1 -.303
Sig. (2-tailed) .249 .046 .091 .143 .718 .458 .889 .622 .000 .435 .509
N 7 7 7 7 7 7 7 7 7 7 7 7
Pearson Correlation
.582 .003 .166 -.610 .746 .640 .803 .723 -.209 .699 -.303 1
Sig. (2-tailed) .170 .996 .722 .146 .054 .122 .030 .066 .652 .081 .509
N 7 7 7 7 7 7 7 7 7 7 7 7
LessThanHS
POC2
Women
OCC_RENTER
MedIllnes
NoVehicle
SocIsol
TotDis
TotChild
OlderAdult
Low_to_No
LEP
West Roxbury Correlations
130
130
Appendix D: Sum and Percentage of Neighborhood Populations by Social Determinants of Vulnerability
Table B1: Sum of Neighborhood Populations
Community2010 Population
2010 Census Tract Count
Social Isolation Disability Children
Older Adults
Low‐to‐No Income
Limited English
Less Than High School
People of Color Women Renters
Medical Illness
No Vehicle
Allston/Brighton 75,009 18 7,745 6,187 4,585 6,087 20,965 27,052 4,243 25,433 35,267 24,175 24,175 12,192
Back Bay/Beacon Hill 22,601 9 3,123 1,044 1,906 2,761 2,555 5,316 231 3,643 11,097 8,766 8,766 7,048
Charlestown 16,439 6 2,502 1,535 3,301 1,811 4,157 5,968 1,296 3,981 7,519 4,314 4,314 1,967
Downtown 30,023 9 4,101 2,602 2,016 4,075 6,783 10,858 2,731 9,424 14,011 11,128 11,128 9,226
East Boston 40,517 14 5,926 5,180 8,665 4,147 13,698 17,845 9,159 25,459 14,874 10,624 10,624 5,754
Fenway/Kenmore 41,788 10 3,473 2,738 646 2,063 11,177 13,240 691 14,449 22,155 13,243 13,243 10,296
Harbor Islands 535 1 ‐ 179 ‐ 12 349 361 66 370 108 ‐ ‐ ‐
Hyde Park 32,317 8 4,904 3,824 6,954 4,174 5,724 9,898 3,147 23,189 13,407 4,930 4,930 2,065
Jamaica Plain 42,160 15 5,801 4,222 6,270 4,094 14,470 18,564 2,785 19,153 19,222 11,718 11,718 6,404
Mattapan 33,682 8 6,230 5,969 9,638 3,869 11,881 15,750 4,910 32,118 14,236 7,649 7,649 4,105
North Dorchester 28,452 8 4,251 3,702 6,389 2,277 10,417 12,694 4,327 18,847 11,489 7,484 7,484 3,868
Roslindale 32,246 10 4,922 4,077 7,146 3,845 6,817 10,662 2,925 16,661 14,029 6,366 6,366 2,659
Roxbury 66,070 20 12,393 10,423 16,689 5,800 27,688 33,488 9,042 59,164 26,950 18,388 18,388 10,651
South Boston 33,674 13 5,006 2,987 4,855 3,233 8,181 11,414 2,497 7,136 15,433 9,719 9,719 4,701
South Dorchester 58,937 14 9,458 8,730 14,589 6,234 16,148 22,382 8,637 43,663 23,756 12,545 12,545 6,398
South End 32,708 10 5,376 4,318 4,908 3,340 11,554 14,894 3,669 16,451 13,652 11,137 11,137 8,089
West Roxbury 30,445 7 4,773 2,984 6,102 5,365 3,495 8,860 1,384 8,143 13,434 4,723 4,723 1,341
Grand Total 617,603 180 89,984 70,701 104,659 63,187 176,059 239,246 61,740 327,284 270,639 166,909 236,938 96,764
131
131
Table B2: Percentages for Neighborhood Populations
Community 2010 Population% 2010 Census
Tract Count %Social Isolation % Disability % Children %
Older Adults %
Low‐to‐No Income % Limited
English %
Less Than High School %
People of Color % Women % Renters %
Medical Illness %
No Vehicle %
Allston/Brighton 12.15% 10.00% 10.33% 8.25% 6.11% 8.12% 27.95% 36.07% 5.66% 33.91% 47.02% 32.23% 32.23% 16.25%
Back Bay/Beacon Hill 3.66% 5.00% 13.82% 4.62% 8.43% 12.22% 11.30% 23.52% 1.02% 16.12% 49.10% 38.79% 38.79% 31.18%
Charlestown 2.66% 3.33% 15.22% 9.34% 20.08% 11.02% 25.29% 36.30% 7.88% 24.22% 45.74% 26.24% 26.24% 11.97%
Downtown 4.86% 5.00% 13.66% 8.67% 6.71% 13.57% 22.59% 36.17% 9.10% 31.39% 46.67% 37.06% 37.06% 30.73%
East Boston 6.56% 7.78% 14.63% 12.78% 21.39% 10.24% 33.81% 44.04% 22.61% 62.84% 36.71% 26.22% 26.22% 14.20%
Fenway/Kenmore 6.77% 5.56% 8.31% 6.55% 1.55% 4.94% 26.75% 31.68% 1.65% 34.58% 53.02% 31.69% 31.69% 24.64%
Harbor Islands 0.09% 0.56% 0.00% 33.46% 0.00% 2.24% 65.23% 67.48% 12.34% 69.16% 20.19% 0.00% 0.00% 0.00%
Hyde Park 5.23% 4.44% 15.17% 11.83% 21.52% 12.92% 17.71% 30.63% 9.74% 71.75% 41.49% 15.26% 15.26% 6.39%
Jamaica Plain 6.83% 8.33% 13.76% 10.01% 14.87% 9.71% 34.32% 44.03% 6.61% 45.43% 45.59% 27.79% 27.79% 15.19%
Mattapan 5.45% 4.44% 18.50% 17.72% 28.61% 11.49% 35.27% 46.76% 14.58% 95.36% 42.27% 22.71% 22.71% 12.19%
North Dorchester 4.61% 4.44% 14.94% 13.01% 22.46% 8.00% 36.61% 44.62% 15.21% 66.24% 40.38% 26.30% 26.30% 13.59%
Roslindale 5.22% 5.56% 15.26% 12.64% 22.16% 11.92% 21.14% 33.06% 9.07% 51.67% 43.51% 19.74% 19.74% 8.25%
Roxbury 10.70% 11.11% 18.76% 15.78% 25.26% 8.78% 41.91% 50.69% 13.69% 89.55% 40.79% 27.83% 27.83% 16.12%
South Boston 5.45% 7.22% 14.87% 8.87% 14.42% 9.60% 24.29% 33.90% 7.42% 21.19% 45.83% 28.86% 28.86% 13.96%
South Dorchester 9.54% 7.78% 16.05% 14.81% 24.75% 10.58% 27.40% 37.98% 14.65% 74.08% 40.31% 21.29% 21.29% 10.86%
South End 5.30% 5.56% 16.44% 13.20% 15.01% 10.21% 35.32% 45.54% 11.22% 50.30% 41.74% 34.05% 34.05% 24.73%
West Roxbury 4.93% 3.89% 15.68% 3.00% 20.04% 17.62% 11.48% 29.10% 4.55% 26.75% 44.13% 15.51% 15.51% 4.40%
Grand Total 100% 100% 15% 11.45% 16.95% 10.23% 28.51% 38.74% 10.00% 52.99% 43.82% 27.03% 38.36% 15.67%
132
132
References
Abkowitz, M. D. (2008). Opertional Risk Management: A Case Study Approach to Effective
Planning and Response. Hoboken, NJ: John Wiley & Sons, Inc.
Assessing Department. (2014). Search FY2014 Real Estate Assessments and Taxes. Retrieved
6/15/2014, from City of Boston http://www.cityofboston.gov/assessing/
Boston Housing Authority. (2014). Development Information. Retrieved 6/15/2014, 2014, from
http://www.bostonhousing.org/housing_dev.html
Brookings-Bern Project on Internal Displacement. (2008). Human Rights and Natural Disasters:
Operational Guidelines and Field Manual on Human Rights Protection in Situations of
Natural Disaster. Washington, D.C.
Chandra, A., Acosta, J., Stern, S., Uscher-Pines, L., Williams, M. V., Yeung, D., . . . Meredith,
L. S. (2011). Building Community Relience to Disasters: A Way Forward to Enhance
National Health Security. Santa Monica: RAND Corporation.
City of Boston. (2014). Chinatown/Leather District. Retrieved 6/15/14, 2014, from
http://www.cityofboston.gov/neighborhoods/chinatownleatherdistrict.asp
Cutter, S. L., & Emrich, C. T. (2006). Moral Hazard, Social Catastrophe: The Changing Face of
Vulnerability along the Hurricane Coasts. Annals of the American Academy of Political
and Social Science, 604(1), 102-112.
Dickson, E., Baker, J. L., & Hoornweg, D. (2012). Urban Risk Assessments: Understanding
Disaster and Climate Risk in Cities: World Bank Publications.
Durant, T. J. (2011). The Utility of Vulnerability and Social Capital Theories in Studying the
Impact of Hurricane Katrina on the Elderly. Journal of Family Issues, 32(10), 1285-1302.
doi: 10.1177/0192513X11412491
133
133
Dwyer, A., Zoppou, C., Nielsen, O., Day, S., & Roberts, S. (2004). Quantifying Social
Vulnerability: A methodology for identifying those at risk to natural hazards. Geoscience
Australia Record, 2004/14.
Federal Emergency Management Agency. (2010a). Comprehensive Preparedness Guide 101:
Developing and Maintaining Emergency Operations Plans, Version 2.0. Washington,
D.C.
Federal Emergency Management Agency. (2010b). Guidance on Planning for Integration of
Functional Needs Support Services in General Population Shelters. Washington, D.C.
Federal Emergency Management Agency. (2011). National Disaster Recovery Framework.
Washington, D.C.: U.S. Department of Homeland Security.
Flanagan, B. E., Gregory, E. W., Hallisey, E. J., Heitgerd, J. L., & Lewis, B. (2011). A Social
Vulnerability Index for Disaster Management. Journal of Homeland Security and
Emergency Management: Vol. 8: Iss. 1, Article 3, 1-21.
Fothergill, A., Maestas, E. G. M., & Darlington, J. D. (1999). Race, Ethnicity and Disasters in
the United States: A Review of the Literature. Disasters, 23(2), 156-173. doi:
10.1111/1467-7717.00111
Fothergill, A., & Peek, L. A. (2004). Poverty and disasters in the United States: A review of
recent sociological findings. Natural Hazards, 32(1), 89-110.
Galea, S., Freudenberg, N., & Vlahov, D. (2005). Cities and population health. Social Science &
Medicine, 60(5), 1017-1033. doi: http://dx.doi.org/10.1016/j.socscimed.2004.06.036
Ganzel, B. L., Morris, P. A., & Wethington, E. (2010). Allostasis and the human brain:
Integrating models of stress from the social and life sciences. Psychol Rev, 117(1), 134-
174. doi: 10.1037/a0017773
134
134
Greenberger, M. (2007). Preparing Vulnerable Populations for Disaster: Inner-city Emergency
Preparedness - Who Should Take the Lead. Journal of Health Care Law, 291-308.
Gustafsson, P. E., San Sebastian, M., Janlert, U., Theorell, T., Westerlund, H., & Hammarström,
A. (2014). Life-Course Accumulation of Neighborhood Disadvantage and Allostatic
Load: Empirical Integration of Three Social Determinants of Health Frameworks.
American Journal of Public Health, 104(5), 904-910.
Hansell, D. (2009). Disaster Case Management: Developing a Comprehensive National Program
Focused on Outcomes, written statement for Ad Hoc Subcommittee on Disaster
Recovery. Paper presented at the Committee on Homeland Security and Governmental
Affairs, US Senate, .
Hawkins, R. L., & Maurer, K. (2009). Bonding, Bridging and Linking: How Social Capital
Operated in New Orleans following Hurricane Katrina. British Journal of Social Work,
40(6), 1777-1793. doi: 10.1093/bjsw/bcp087
Intergovernmental Panel on Climate Change. (2012). Managing the Risks of Extreme Events and
Disasters to Advance Climate Change Adaptation. In C. Field, V. Barros, T. Stocker, Q.
Dahe, D. J. Dodden, K. L. Ebi, M. D. Mastrandrea, K. Mach, G.-K. Plattner, S. K. Allen,
M. Tignor & P. Midgley (Eds.), Special Report of the Intergovernmental Panel on
Climate Change. New York.
Isaranuwatchai, W., Coyte, P. C., McKenzie, K., & Noh, S. (2013). Impact of the 2004 Tsunami
on Self-Reported Physical Health in Thailand for the Subsequent 2 Years. American
Journal of Public Health, 103(11), 2063-2070.
Issenberg, S. (2012). The victory lab: The secret science of winning campaigns: Random House
LLC.
135
135
Kailes, J. I., & Enders, A. (2007). Moving Beyond "Special Needs". Journal of Disability Policy
Studies, 17(4), 230-237.
Klinenberg, E. (1999). Denaturalizing disaster: A social autopsy of the 1995 Chicago Heat
Wave. Berkeley: University of California, Berkeley.
Klinenberg, E. (2002). Heat wave: a social autopsy of disaster in Chicago. Chicago: The
University of Chicago Press.
Knobel, C. (2013). Social network analysis as an augmentation of qualitative research. Paper
presented at the Collaboration Technologies and Systems (CTS), 2013 International
Conference on.
Levac, J., Toal-Sullivan, D., & O'Sullivan, T. L. (2012). Household emergency preparedness: a
literature review. J Community Health, 37(3), 725-733. doi: 10.1007/s10900-011-9488-x
Logan, J. G., & Barksdale, D. J. (2008). Allostasis and allostatic load: expanding the discourse
on stress and cardiovascular disease. Journal of Clinical Nursing, 17(7b), 201-208. doi:
10.1111/j.1365-2702.2008.02347.x
MassHousing. (2014). Housing List. Boston, MA.
Mileti, D. S. (1999). Disasters by design: A reassessment of natural hazards in the United States.
Washington, D.C.: National Academies Press.
Milligan & Company, L., ,. (2007). Transportation Equity in Emergencies: A Review of the
Practices of State Departments of Transportation, Metropolitan Planning Organizations,
and Transit Agencies in 20 Metropolitan Areas (Vol. 2014).
Mohaupt, S. (2009). Review Article: Resilience and Social Exclusion. Social Policy and Society,
8(1), 63-71. doi: 10.1017/S1474746408004594
136
136
Molin Valdes, H., Rego, A., Scott, J., & Aguayo, J. V. (2012). How to Make Cities More
Resilient: A Handbook for Local Government Leaders. In P. Bittner (Ed.). Geneva.
Morrow, B. H. (1999). Identifying and Mapping Community Vulnerability. Disasters, 23(1), 1-
18.
Muñiz, B. (2006). In the eye of the storm: how the government and private response to Hurricane
Katrina failed Latinos. Washington, DC.
Norman-Major, K. (2011). Balancing the Four Es; or Can We Achieve Equity for Social Equity
in Public Administration? Journal of Public Affairs Education, 17(2), 233-252.
Pantell, M., Rehkopf, D., Jutte, D., Syme, S. L., Balmes, J., & Adler, N. (2013). Social isolation:
a predictor of mortality comparable to traditional clinical risk factors. American Journal
of Public Health, 103(11), 2056-2062. doi: 10.2105/AJPH.2013.301261
Paulison, R. D. (2005). Civil Rights Program. Washington, D.C.: U.S. Department of Homeland
Security.
Pelling, M. (2003). Vulnerability in Cities. London: Earthscan Publications.
Rosso, P. (2012). Northampton Street project gets the OK from city, Boston Globe. Retrieved
from
http://www.boston.com/yourtown/news/roxbury/2012/09/northampton_street_project_get
.html
Seeman, T. E. (1996). Social ties and health: The benefits of social integration. Annals of
epidemiology, 6(5), 442-451.
Sherry, J. M., Nishamarie, & Harkins, J. D., Anne Marie. (2011). Leveling the emergency
preparedness playing field. Journal of Emergency Management, 9(6), 11-16. doi:
10.5055/jem.2011.0075
137
137
Shi, L., & Stevens, G. D. (2010). Vulnerable populations in the United States (Vol. 23): John
Wiley & Sons.
Swiss Reinsurance Company Ltd. (2013). Mind the Risk: A global ranking of cities under threat
from natural disasters. Zurich.
U.S. Census Bureau. (2012). 2012 ACS 5-Year Estimates. Retrieved from:
http://factfinder2.census.gov/faces/nav/jsf/pages/download_center.xhtml
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_12_5Y
R_DP04&prodType=table
U.S. Census Bureau (Producer). (2013, August 11). Growth in Urban Population Outpaces Rest
of Nation, Census Bureau Reports. Census.gov. Retrieved from
http://www.census.gov/newsroom/releases/archives/2010_census/cb12-50.html
U.S. Centers for Disease Control and Prevention. (2010). Public Health Workbook: To Define,
Locate, and Reach Special, Vulnerable, and At-Risk Populations in an Emergency.
Washington, D.C.
U.S. Centers for Disease Control and Prevention (CDC). (2011). Public Health Preparedness
Capabilities: National Standards for State and Local Planning. Atlanta.
U.S. Department of Health and Human Services. (2009). National Health Security Strategy of
the United States of America. Washington, D.C.
U.S. Department of Homeland Security. (2010a). Comprehensive Preparedness Guide 101:
Developing and Maintaining Emergency Operations Plans. Washington, D.C.
U.S. Department of Homeland Security. (2011). National Preparedness Goal, First Edition.
Washington, D.C.
138
138
U.S. Department of Homeland Security. (2013a). National Mitigation Framework. Washington,
D.C.
U.S. Department of Homeland Security. (2013b). National Response Framework (Second ed.).
Washington, D.C.
U.S. Department of Homeland Security. (2013c). NIPP 2013: Partnering for Critical
Infrastructure Security and Resilience. Washington, D.C.
U.S. Department of Homeland Security. (2014). 2014 Quadrennial Homeland Security Review.
Washington, D.C.: U.S. Department of Homeland Security.
United Nations Office for Disaster Reduction. (2005). Hyogo Framework for Action 2005-2015:
Building the Resilience of Nations and Communities to Disasters. Geneva.
Wisner, B., & Handmer, J. (1999). Hazards, globalisation, and sustainability. Paper presented at
the Development in Practice.