&EPA
United States
Environmental Protection
Agency
EPA/600/R-16/178 | September 2016 | vwvw.epa.gov/research
Sustainability at the Community Level:
Searching for Common Ground as a
Part of a National Strategy for Decision
Support
National Health and Environmental Effects Research Laboratory
Office of Research and Development
-------
EPA/600/R-16/178
September 2016
Sustainability at the Community Level:
Searching for Common Ground
as a Part of a National Strategy
for Decision Support
by
Richard S. Fulford
Marc Russell
Jim Harvey
Matthew C. Harwell
Gulf Ecology Division
National Health and Environmental Effects Research Laboratory
Gulf Breeze, FL 32561
National Health and Environmental Effects Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
-------
Notice/disclaimer statement
The U.S. Environmental Protection Agency through its Office of Research and Development (ORD)
funded and collaborated in the research described herein. This document has been reviewed by the U.S.
Environmental Protection Agency, Office of Research and Development, and approved for publication.
Any mention of trade names, products, or services does not imply an endorsement by the U.S.
Government or the U.S. Environmental Protection Agency.
This is a contribution to the EPA ORD Sustainable and Healthy Communities Research Program.
The appropriate citation for this report is:
Fulford, R.S., M. Russell, J. Harvey, and M.C. Harwell. 2016. Sustainability at the Community Level:
Searching for Common Ground as a Part of a National Strategy for Decision Support. U.S.
Environmental Protection Agency, Gulf Breeze, FL, EPA/600/R-16/178.
-------
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the Nation's
land, air, and water resources. Under a mandate of national environmental laws, the Agency strives to
formulate and implement actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. To meet this mandate, EPA's research program is
providing data and technical support for solving environmental problems today and building a science
knowledge base necessary to manage our ecological resources wisely, understand how pollutants affect
our health, and prevent or reduce environmental risks in the future.
The National Health and Environmental Effects Research Laboratory (NHEERL) within the Office of
Research and Development (ORD) is the Agency's center for investigation of technological and
management approaches for reducing anthropogenic effects that threaten human health and the
environment. The focus of the Laboratory's research program is on models, tools, and approaches for
identification, understanding, measurement, and prevention of anthropogenic effects to air, land, water,
and subsurface resources; protection of air, water, sediments, and ground water; and restoration of
ecosystems. NHEERL collaborates with both public and private sector partners to foster new tools and
approaches that address existing issues and anticipate emerging problems. NHEERL's research provides
solutions to environmental problems by: developing and promoting technologies that protect and
improve the environment; advancing scientific and engineering information to support regulatory and
policy decisions; and providing the technical support and information transfer to ensure implementation
of environmental regulations and strategies at the national, state, and community levels.
This report describes NHEERL's investment into local decision support through examination of
community characteristics that are associated with decision priorities, local available resources, and
meaningful measures of human well-being. This comparison directly supports NHEERL's research
priorities through the examination of transferability and utility of existing tools and approaches for
decision support at the community level.
William Benson, Acting Laboratory Director
National Health and Environmental Effects Research Laboratory
-------
Table of Contents
Notice/disclaimer statement ii
Foreword iii
Figures vi
Tables viii
Acknowledgments xi
Acronyms and abbreviations xii
Executive summary xiii
1 Introduction 1
1.1 Literature cited 4
2 Community classification system 5
2.1 Methods 5
2.1.1 Input data 5
2.1.2 Data analysis 7
2.2 Results 8
2.2.1 Cluster analysis 8
2.2.2 Group description 14
2.3 Discussion 14
2.4 Literature cited 16
3 Human well-being across community types 17
3.1 Methods 18
3.1.1 Classifying communities 18
3.1.2 Measuring human well-being 18
3.1.3 Stakeholder-derived weightings for domains of human well-being index 19
3.2 Results 20
3.2.1 Human well-being and community type 20
3.2.2 Weighted vs. unweighted human well-being index 22
3.3 Discussion 25
3.4 Conclusions 28
3.5 Literature cited 29
4 Stakeholder-derived priorities 32
4.1 Direct stakeholder engagement 32
4.1.1 Introduction 32
iv
-------
4.1.2 Overall goals 35
4.1.3 Methods 35
4.1.4 Results 42
4.1.5 Discussion 54
4.2 Keyword-based analysis of community planning documents 60
4.2.1 Introduction 60
4.2.2 Methods 60
4.2.3 Results 63
4.2.4 Discussion 71
4.3 Engagement conclusions 73
4.4 Literature cited 76
5 Ecosystem goods and services 77
5.1 Introduction 77
5.2 Methods 77
5.2.1 Usable water and stable climate 78
5.2.2 Usable air 79
5.2.3 Flood protection 80
5.3 Results 81
5.4 Discussion 99
5.5 Literature cited 101
6 Synthesis 103
6.1 Literature cited 105
Appendix A: Sample community workshop agenda 106
Appendix B: List of goals 107
Appendix C: Counties in keyword search 110
Appendix D: Keyword lists 111
Appendix E: R code for keyword analysis annotated 126
Appendix F: References for denitrification rates and carbon burial rates 131
Appendix G: NLCD category descriptions 132
Appendix H: Percentage of each soil type by NLCD land cover categories for the four
counties 133
Appendix I: Curve numbers 134
V
-------
Figures
Figure 2.1 Map indicating the counties in the conterminous United States defined as coastal
counties for this analysis (black outline) 7
Figure 2.2 Scree plot indicating the distribution of 663 coastal counties with respect to the first
two principal components 8
Figure 2.3 Tree plot indicating the eight cluster groups delineated for this analysis (cut point
indicated in red) 9
Figure 2.4a Summary plots showing the distribution of mean population density (mi-2) for
coastal counties by Group (A) 11
Figure 2.4b Summary plots showing the mean (SD) LQ scores for all three categories by
Group (B) 11
Figure 2.5 Map of coastal counties with group membership indicated by color 12
Figure 3.1 Summary of the human well-being index (HWBI) scores by community classification
group 21
Figure 3.2 Comparison of mean weighted (wted) and unweighted (unwted) values of the
human well-being index (HWBI) among the nine communities included in the study 25
Figure 4.1 Communities participating in community engagement for sustainability workshops.
33
Figure 4.2 Relative frequency of reporting for goal categories identified during mapping
exercises 45
Figure 4.3 Relative frequency of goal categories receiving dot votes during workshop group
voting exercise 46
Figure 4.4 Likelihood a goal category was identified as top three priority during a workshop in
an individual ranking exercise 48
Figure 4.5 Comparison of workshop outcomes across the three weighting exercises 50
Figure 4.6 Relationship between four types of EPA-identified community sustainability
indicators 55
Figure 4.7 Summary of normalized differences between keyword and manual reads of selected
test documents 64
Figure 4.8 Summary of the median across all documents analyzed (58) for normalized hits per
domain of the human well-being index 66
Figure 4.9 Scree plot summarizing results of a principal components analysis (PC) of
normalized hits per domain of the human well-being index among with the full suite of
independent variables considered (See Chapter 4.2.2 for details) 71
Figure 4.10. Summary comparison of combined results from workshops (Chapter 4.1) and
keyword analysis (Chapter 4.2) in four focal communities 74
Figure 5.1 Maps showing National Land Cover Dataset (NLCD) coverage (See Table 5.2) for:
(A) Escambia County, FL; (B) Indian River County, FL; (C) Lafourche Parish, LA; and (D)
St. Landry Parish, LA 82
Figure 5.2 Impervious land cover as reported by the NLCD 2011 Percent Developed
Imperviousness data (See Table 5.3) for: (A) Escambia County, FL; (B) Indian River
County, FL; (C) Lafourche Parish, LA; and (D) St. Landry Parish, LA 87
vi
-------
Figure 5.3 Canopy cover as reported by the NLCD 2011 USGS Tree Canopy cartographic data
(See Table 5.4) for: (A) Escambia County, FL; (B) Indian River County, FL; (C) Lafourche
Parish, LA; and (D) St. Landry Parish, LA 88
Figure 5.4 Denitrification rates averaged from literature review (See Table 5.5) for each NLCD
category for: (A) Escambia County, FL; (B) Indian River County, FL; (C) Lafourche Parish,
LA; and (D) St. Landry Parish, LA 89
Figure 5.5 Carbon burial rate averaged from literature review (See Table 5.6) for each NLCD
category for: (A) Escambia County, FL; (B) Indian River County, FL; (C) Lafourche Parish,
LA; and (D) St. Landry Parish, LA 90
Figure 5.6 Usable water value as calculated using average denitrification rates for each NLCD
category (See Table 5.7) for: (A) Escambia County, FL; (B) Indian River County, FL; (C)
Lafourche Parish, LA; and (D) St. Landry Parish, LA 91
Figure 5.7 Usable air value as calculated using the average canopy cover percentage for each
NLCD category (See Table 5.8) for: (A) Escambia County, FL; (B) Indian River County, FL;
(C) Lafourche Parish, LA; and (D) St. Landry Parish, LA 92
Figure 5.8 Stable climate value as calculated using average carbon burial rates for each NLCD
category (See Table 5.9) for: (A) Escambia County, FL; (B) Indian River County, FL; (C)
Lafourche Parish, LA; and (D) St. Landry Parish, LA 93
Figure 5.9 Flood protection value as calculated using averaged curve numbers for each NLCD
category (See Table 5.10) for: (A) Escambia County, FL; (B) Indian River County, FL; (C)
Lafourche Parish, LA; and (D) St. Landry Parish, LA 94
-------
Tables
Table 2.1 Summary of community classification system data and results by Group including a
qualitative description of each Group based on the data (Description), number of counties
included in each Group (n), States with counties in each Group (flagged if only one
county), highest and lowest Lifemode (LM) category for each county by percentage, mean
(SD) LQ scores for each county by employment category (Throughput, Local Dependence
and Service), and the maximum percentage among the 85 Ecoregion categories indicating
the evenness of coverage across the eight Groups 13
Table 3.1 Relative frequency of goal categories receiving dot votes during workshop group
voting exercise 23
Table 3.2 Summary of weighted (Wt) and unweighted (Uwt) scores for the human well-being
index (HWBI) 24
Table 4.1 Workshop communities and central issues 33
Table 4.2 Key design elements and relationships to workshop objectives 33
Table 4.3 Workshop participation for major interest groups by community 39
Table 4.4 Characteristics of four study communities 39
Table 4.5 Relative frequency of reporting for goal categories identified during mapping
exercises 45
Table 4.6 Relative frequency of goal categories receiving dot votes during workshop group
voting exercise 46
Table 4.7 Likelihood a goal category was identified in top three during a workshop in an
individual ranking exercise 47
Table 4.8 Summary of univariate cluster analyses between communities 52
Table 4.9 Summary of associations' analysis across communities 53
Table 4.10 Strategic planning documents used for the validation of the keyword list used in this
analysis 61
Table 4.11 Summary of line by line matches between keyword and manual reads of test
documents organized by HWBI domain 65
Table 4.12 Summary of normalized keyword hits organized by domains of human well-being
index 66
Table 4.13 Summary of mean normalized keyword hits for the domains of human well-being
index organized by U.S. state 67
Table 4.14 Summary of mean normalized keyword hits for the domains of human well-being
index organized by coastal CCS groups 68
Table 4.15 Summary of mean normalized keyword hits for the domains of human well-being
index organized by proportion of adults citizens with either a high school diploma or a
bachelor's degree in 2000 (U.S. Census Data; accessed 14 September 2016) 68
Table 4.16 Summary of mean normalized keyword hits for the domains of human well-being
index organized by median income level, median age, and population size in 2000 (U.S.
Census Data; accessed 14 September 2016) 69
viii
-------
Table 4.17 Summary of mean normalized keyword hits for the domains of human well-being
index organized by proportion of community self-reporting in three ethnic groups in 2000
(U.S. Census Data; accessed 14 September 2016) 70
Table 5.1 Summary of external data used in calculation of ecosystem goods and service
delivery 78
Table 5.2 Total area and percentage of areal coverage of NLCD land cover categories for four
counties 83
Table 5.3 Percentage of impervious land cover in each of four counties 84
Table 5.4 Percentage of areal canopy cover by land cover category for four counties 84
Table 5.5 Denitrification rates (g N/m2/yr) by NLCD category for each county 85
Table 5.6 Carbon burial rates (g C/m2/yr) for each NLCD category in four counties 85
Table 5.7 Value ($/ha/yr) of maintaining water quality via natural denitrification by land cover
categories for four counties 95
Table 5.8 Value ($/ha/yr) of maintaining air quality via natural carbon processing in the canopy
cover 95
Table 5.9 Value ($/ha/yr) of maintaining stable climate via natural carbon burial 96
Table 5.10 Value ($/ha/yr) of maintaining flood protection based on soil characteristics by land
cover categories for four counties 97
Table 5.11 Summary table of total value per hectare for select ecosystem goods and services
for four counties 98
Table 5.12 Summary table of total value per capita per year for select ecosystem goods and
services for four counties 98
Table B.1 List of goals 107
Table B.2 Detailed explanation of goals 108
Table D.1 Keyword list for connection to nature 111
Table D.2 Keyword list for cultural fulfillment 112
Table D.3 Keyword list for education, basic knowledge 113
Table D.4 Keyword list for education, participation 114
Table D.5 Keyword list for education, development 114
Table D.6 Keyword list for health, healthcare 115
Table D.7 Keyword list for health, personal well-being 115
Table D.8 Keyword list for health, physical, and mental health conditions 116
Table D.9 Keyword list for health, life expectancy 116
Table D.10 Keyword list for health, lifestyle 117
Table D.11 Keyword list for leisure time, activity participation 118
Table D.12 Keyword list for leisure time, retired seniors 119
Table D.13 Keyword list for leisure time, time spent 119
Table D.14 Keyword list for leisure time, working age adults 119
Table D.15 Keyword list for living standards, basic necessities 120
Table D.16 Keyword list for living standards, income 121
Table D.17 Keyword list for living standards, wealth 121
-------
Table D.18 Keyword list for living standards, work 121
Table D.19 Keyword list for safety and security, actual safety 122
Table D.20 Keyword list for safety and security, perceived safety 122
Table D.21 Keyword list for social cohesion, attitude towards community 123
Table D.22 Keyword list for social cohesion, democratic engagement 123
Table D.23 Keyword list for social cohesion, family bonding 124
Table D.24 Keyword list for social cohesion, social engagement 124
Table D.25 Keyword list for social cohesion, social support 125
Table F.1 References for denitrification rates 131
Table F.2 References for carbon burial rates 131
Table G.1 NLCD category descriptions 132
Table H.1 Percentage of each soil type by NLCD land cover categories for the four counties
133
Table 1.1 Curve number by land cover category and soil type 134
Table 1.2 Weighted curve number by land cover category for four counties assigned based on
the soil hydrogroup of each NLCD class 135
X
-------
Acknowledgments
This synthesis report was prepared by the U.S. Environmental Protection Agency (EPA), Office of
Research and Development (ORD), National Health and Environmental Effects Research Laboratory
(NHEERL), Gulf Ecology Division (GED). Editorial assistance was provided by Chloe Jackson and
Kate Murphy. The following task members provided written materials and technical information in the
preparation of this document.
• Richard S. Fulford, Office of Research and Development
• Lisa M. Smith, Office of Research and Development
• Marc Russell, Office of Research and Development
• Susan Yee, Office of Research and Development
• Ian Krauss, former Student Services Contractor
• Kate Murphy, Student Services Contractor
The following external contractors provided written materials and technical information in the
preparation of this document under contract EP-W-11-010.
• Dr. Bill Michaud, SRI International, Inc.
• Theresa Nishimoto, SRI International, Inc.
-------
Acronyms and abbreviations
ACS
American Community Survey
ANOVA
Analysis of Variance
CBDS
Community-Based Decision Support
CCS
Community Classification System
CIC
Community Indicators Consortium
CN
Curve Number
CR
Caribbean Regions
EAB
Ecosystems Assessment Branch
EGS
Ecosystem Goods and Services
EPA
Environmental Protection Agency
GED
Gulf Ecology Division
GIS
Geographic Information System
GOM
Gulf of Mexico
HWB
Human Weil-Being
HWBI
Human Weil-Being Index
ICMA
International City/County Management Association
LM
Lifemode
LQ
Location Quotient
LULC
Land Use Land Cover
MUKEY
Map Unit Key
NAICS
North American Industry Classification System
NHEERL
National Health and Environmental Effects Research Laboratory
NLCD
National Land Cover Dataset
NOAA
National Oceanic and Atmospheric Administration
OECD
Organization for Economic Co-Operation and Development
ORD
Office of Research and Development
RESES
Regional Sustainable Environmental Science
RIVs
Relative Importance Values
SHC
Sustainable and Healthy Communities
SSURGO
Soil Survey Geographic Database
USDA
United States Department of Agriculture
USFS
United States Forest Service
USGS
United States Geological Survey
-------
Executive summary
The Sustainable and Healthy Communities (SHC) research program is intended to support resource
sustainability and decision making at the community level. Sustainability is defined as the ability of a
community to meet present needs without compromising the ability of society and the environment to
meet the economic, social, and environmental needs of future generations. The USEPA and its partners
seek a national strategy that maximizes impacts by identifying common ground among communities that
can inform the decision process. In this report, communities are compared based on four distinct metrics
(community type; human well-being index; stakeholder priorities; and availability of ecosystem goods
and services) with the purpose of seeking common ground for defining and measuring sustainability at
the local scale. Overlying this comparison is the question of the usefulness of a community classification
system (CCS) for generalizing the findings to new communities.
Community type was found to be informative regarding the relative importance of elements of well-
being. Two major delineations of community type are considered here. First is geographic, or simply
asking if a place defines how communities measure well-being. The second was the CCS described in
debate in Chapter 2. We then examine whether values of a specific measure of well-being, the human
well-being index (HWBI), differ either geographically or by community type. Stakeholder priorities are
then examined in Chapter 4, with two methods, both involving elements of the HWBI. The objective
was to link stakeholder priorities to HWBI and look for differences in these priorities among
communities. Finally, we examined if available ecosystem resources differ either geographically or by
community type and provide some recommendations for using all of the information as a part of a
national strategy for classifying communities in support of decision making for sustainability.
The analysis in Chapters 2 and 3 involves the description of a CCS and the amount of information
regarding human well-being (HWB) contained in the CCS. This is important because community
decision makers may use the CCS to help identify baseline well-being values from which to assess the
impact of decisions as shifts in community-specific HWB. Measures of community-specific HWB also
allow communities to restore, achieve and sustain what matters most to them in terms of human well-
being. The environmental components (e.g., ecoregion) of the CCS were less informative about
community type than the economic and social components (i.e., Lifemode and Location Quotient), yet
the differences in community type were strongly driven by economic and social dependence on local
environmental resources either through employment or through land use. This finding points to a clear
link between environmental service flows and HWB.
The approach of setting local HWB reference points based on community classification assumes that
common ground is important for describing community priorities. The limitations of this approach are
that specific factors important to individual communities are not considered and are likely to change in
importance across communities and at different spatial scales than considered here. Decision makers
wishing to set reference points for HWB will need to consider the consistency of the group assignments
to their situation, but in cases where this is an effective approach, much will be gained by allowing
similar communities to compare their HWB values.
The community classification system developed during this study was also intended to inform decision
makers about a community's priorities. The association of these priorities with human well-being is a
tool for informing decision makers about sustainable decision outcomes in a community-specific
context. Stakeholder engagement is an important tool for understanding the priorities of a community. In
-------
Chapter 4, two methods for stakeholder engagement were explored with the HWBI as an engagement
framework in each case. In Chapter 4.1, a workshop approach is described based on structured decision
making (Structured Decision Making; accessed 14 September 2016), while in Chapter 4.2 an automated
analysis of strategic planning documents is described based on keyword counting method. Key
differences were observed in the outcomes of these two methods. The workshop method generated more
diverse findings that nonetheless consistently reported high importance in the domains of Education and
Social Cohesion. In contrast, the keyword method was always dominated by Living Standards, which is
the primary economic domain of HWBI. In terms of meaning, the keyword results are based on strategic
planning, which is predictably action-focused and heavily weighted to economic aspects of a
community's well-being. In contrast, workshop results show a broader influence and this is likely the
result of facilitation and the separation of community priorities from a particular action (Chapter 4.1).
The findings of the keyword analysis can be thought of as hierarchical with the secondary outcomes
being more similar to workshop outcomes. There is, thus, strong support for the complementarity of the
two methods. Consistent results across the two approaches provide good support for the complementary
nature of the data and the value of applying both methods simultaneously to identify community
priorities.
The stakeholder workshops held as a part of this study generated important insights into the nature and
hierarchical structure of core community values and implications for indices of sustainability.
Communities participating in the workshops demonstrated an innate capacity for systems thinking, and
this suggests that in the context of community decisions and action, values associated with the most
fundamental aspects of well-being could be the highest priorities. Practical sustainability indices will
need to be adaptable to changes in a way that measures and emphasizes core values that remain high
priorities over time and values associated more immediate priorities. The workshops also afforded an
opportunity to explore the elements of the HWBI, particularly the relative importance values (RIV), the
factors used to weight different domain scores to derive element scores (e.g., economic well-being) and
an overall HWBI value. Workshop findings suggest that, from a community perspective, a set of indices
or indicators, rather than aggregated indices, may be more responsive to community needs. The RIVs
could also change over time. This suggests the need to periodically update RIVs.
There is always a question regarding the within-community generalizability of workshop findings with
respect to core community values. In this case, there were promising linear associations between the
priority placed on Education (based on mapping and ranking exercises) and the shares of households in
the four participating communities with children and youth; as well as between the priority placed on
Health (based on mapping and/or dot voting), a critical factor affecting household expenses, and the
unemployment rates and percentages of owner-occupied and renter households that spend 35% or more
of their income on housing costs in the four communities. The analysis of workshop data also revealed
no significant bias in terms of higher prioritization of goals and values that are most closely aligned with
the central issues. The ability to generalize the results of community engagement workshops to the
whole community can be improved by holding multiple workshops at different times of the day, week
and year and by holding workshops in different forums.
Similar to workshops, keyword analysis of strategic planning documents shows great promise as a
contributing method for clarifying the long-term priorities of stakeholders. Clarifying community
priorities from document analysis is limited by the scope of the document, as well as the level to which
the document reflects community input rather than the input of elected officials or hired external experts.
Yet, these issues can largely be minimized by appropriate document selection. A key consistency among
communities in this analysis was the importance of quality of life metrics to stakeholder priorities.
Across communities and community types the consistently dominant domains, in terms of total number
-------
of keyword hits, were Living Standards followed by Safety and Security followed closely by Social
Cohesion and Leisure Time. An interest in quality of life seems to be a common community attribute,
which is not surprising. The consistent low scores for either Connection to Nature and Health were
surprising, but suggest these are not community-level priorities but may be important at a different scale
(e.g., personal/family). For instance, even in cases where an action may directly benefit human health
(e.g., investment in hospitals) the community-scale priority for the action may not be directly tied to
health, but rather to ancillary benefits more aligned with community-scale priorities such as job creation,
reductions in burden on public services, or community reputation. These differences can be important to
setting measures of success at the appropriate scale. It is also important to understand if these results
differ among community types.
The dominant delineations for stakeholder priorities at the community level were between states and
CCS groups. States differed most for Safety and Social Cohesion, while CCS groups differed most in
Living Standards and Leisure time. The less commonly mentioned domains such as Connection to
Nature were more important in specific categories such as median age and ethnic composition of the
community. The value of understanding these differences among groups is to identify the domains of
human well-being for which the CCS or geographic delineations are the most informative. These most
informative differences lie on a gradient from an emphasis on Safety and Living Standards on one end to
an emphasis on Leisure Time and Social Cohesion on the other end. This gradient is also consistent with
an urban to rural gradient in that it is directly related to population size, and demographics as 'ruralness'
tends to be related to an increased emphasis on social connectivity. As communities become more
urban, more diverse, or less dependent on local natural resources they seem to prioritize Safety, Living
Standards and Connection to Nature; and reduce priorities for Social Cohesion and Education. The most
informative delineation of keyword data at the community scale is for CCS groups followed by state
differences, but other delineations become more important at smaller scales within the community.
Domains such as Connection to Nature and Education do not parse out very well at the community
scale, as indicated by the lack of difference among communities for these domains, and the lack of
information about them contained in categories such as CCS and geography. Nonetheless, they can be
quite important in driving individual priorities and so have a collective influence at the community level
not well captured by review of community planning documents. As such, it is not advised that any
conclusions can be drawn about community priorities for these domains with a keyword-based method.
These findings strongly suggest that keyword analysis combined with a CCS based comparison can be
very informative regarding differences in the relative importance of community-scale priorities such as
Social Cohesion, Living Standards, Leisure Time, and Safety.
Beyond the specifics, it is evident that communities differ in how they rank and prioritize the domains of
human well-being and these differences are predictable based on community type. This indicates the
value of community delineations for informing the decision process. However, it also indicates that
measures of success can only be partially generalized and the very definition of human well-being may
differ among community types. Such differences must be kept in mind when comparing the objective
well-being across communities, particularly along the urban to rural gradient. Therefore, use of this
technique in the future should focus on improving the understanding of how community type may
inform differences in the importance of the domains of human well-being that can be used to both
develop and assess decision options at the community level.
Overall, stakeholder priorities were more consistent across communities than across community types.
For both analytical methods, community type was most informative about the relative importance of low
scoring domains of HWBI such as Connection to Nature and Cultural Fulfillment. This is important
information for scoring HWBI and will be used to explore relative weighting within HWBI, but the
XV
-------
dominance of Living Standards, Safety, Education, and Social Cohesion was consistent in both
stakeholder engagement approaches and so seems robust to categorization. Community-specific
deviations were more evident. However, community-level differences are to be expected and the
overarching consistency of multiple domains across communities suggests important common themes
that should be explored for their value in informing and measuring the success of community level
decision support.
Alongside delineation of HWB and stakeholder priorities are measurable differences among
communities in the production and availability of ecosystem goods and services (EGS) that support
decision making (Smith et al. 2013). Ecosystem goods and services represent a community's ties to the
local environment and as such contribute to economic stability, sense of place, and community identity
(Smith et al. 2013). In Chapter 5, we examine how well two delineations of communities (i.e., CCS,
state) inform about community priorities and therefore aid efforts to inform the local decision process.
The largest difference in EGS value between groups was for CCS with the exception of useable water,
which differed more by U.S. state. Urban (CCS type 1) communities in both LA and FL had higher
specific value for usable air and flood protection, while more rural (CCS type 3) communities were
consistently lower in total area of both developed land and forest, and highest in wetlands, the latter
which provide higher denitrification but the former provide more carbon burial and water retention
during flood events. These differences suggest tradeoffs exist between EGS categories in terms of
benefits to humans. In the abstract it seems plausible that flood protection, high denitrification, and high
carbon burial could co-exist at the spatial scale of this analysis (10-100 km), but in practice different
land cover types contributed to each and that land cover types were both distributed differently and
affected differently by human development linked to changes in impervious surface and canopy cover.
Carbon burial, which contributes to a more stable climate, and flood protection are clearly affected by
development and the level of urbanization in a community. Denitrification, which contributes to clean
water, differed more by state than CCS group indicating a lower impact from development but a
stronger regional influence. These realized tradeoffs are important in that they can help clarify
differences in the impacts of development likely to affect decision outcomes. These trade-offs also
support the conclusion that local priorities for sustainability may differ based on the existing high value
services they need to sustain and/or improve and thus CCS groups can help inform the prioritization
process. This conclusion is tied to the notion that spatial demand for ecosystem services is the reciprocal
of spatial supply.
An important overarching question for this report is how the USEPA and its partners should make use of
CCS and HWBI as a part of a national strategy for local decision support. Community-based decision
support is a national scale issue in that the collective impacts of multiple local decisions can have large
and pervasive results on resource sustainability particularly in coastal areas. Central to the question of
national- or regional-scale community decision support is the balance between treating all communities
the same or focusing on the unique issues of each individual community. Treating all communities the
same in the design of metrics and tools is risky because it allows for avoidable variability in community
characteristics to bias the evaluation of metrics and tools, and the resulting tools may be viewed as
'externally driven', which limits the acceptability of the support by community stakeholders. In contrast,
treating each community as totally unique is inefficient and ignores potentially valuable commonalities.
A key focus of this work has been to consider how this balance should be struck in practice, and the
outcome is that a CCS can be a valuable way to approach the issue. The CCS examined in this report
shows promise as a generalizing tool for decision support and more importantly linking it to HWBI
allows for structured local input 'what matters', so that the approach is transferable and adaptable as
needed. Yet, well-being is a moving target and measuring human benefit is tied to tradeoffs in access to
-------
natural resources and most importantly changes across the rural to urban gradient. Therefore, a balance
is proposed between subjective and objective criteria in measuring well-being sustainability at the local
level that may be best achieved through use of the weighted HWBI examined in Chapter 3. Exploration
of methods for effectively applying HWBI/CCS at the community level is an important research
question. The collective outcome of this report strongly supports exploration of a balanced approach for
local decision support that begins with identification of community type and the calculation of weighted
HWBI. Community-level decision support is a national scale issue and should be approached with a
coherent national strategy by seeking common tools to inform similar decisions across multiple
communities. Doing so will maximize the impact of EPA-led efforts and can result in a more effective
and accepted measure of community sustainability.
-------
1 Introduction
The Sustainable and Healthy Communities (SHC)
national research program is intended to support
sustainable decision making at the community level.
The goal of this SHC study was to provide
scientifically sound and user-friendly guidance on the
sustainability of current and proposed community
actions to stakeholder groups, including planners,
decision makers, and the general public. Decision
support at the community level can have far-reaching
implications for environmental quality and human
health and well-being. At the local level, community
decision making impacts important issues like changes
in land use-land cover, which in turn affects air and
water quality (Abdul-Aziz and Al-Amin 2016, Fruet et
al. 2016). More importantly, at the regional- and
national-scale decisions made by multiple communities
can have cumulative effects that are more far-reaching
than the boundaries of the communities that make them
(Tanaka et al. 2016). Community-based decision
support is a national issue that requires common ground for advising all communities about the
implications of their actions. Yet, all communities have important differences in composition, priorities,
and issues that create challenges for forging a coherent national strategy for decision support. Simply
"recreating the wheel" in each community is costly and inefficient, and it is the goal in this report to
explore the similarities among communities in key areas to produce a roadmap for comparability useful
for informing local decision support in environmental planning and protection.
In making comparisons it is important to have a clear understanding of what defines a community and
what communities mean by sustainability. Both terms have been widely used and are not consistently or
easily defined (Portney 2013). For the purposes of this report, a community is defined as any area under
the authority of municipal decision making. This definition is focused on the spatial scope of the
decision maker rather than that of the effects of the decision, which may be much broader (e.g.,
watershed). This choice of scope is purposeful and is tied to the definition of sustainability.
Sustainability is broadly defined as the ability of a community to meet present needs without
compromising the ability of society and the environment to meet the economic, social, and
environmental needs of future generations, but here we focus on actions that support the long-term
provision of human well-being. Human well-being will be more formally defined in a later chapter but
generally represents the collective benefit to community stakeholders from social, economic, and
environmental capital. In other words, a sustainable outcome of community decision making is one that
conserves or restores capital services to community stakeholders.
Community sustainability defined in this way is highly dependent on community characteristics and the
definition of services provided by available social, economic, and environmental capital. Communities
differ in both their available resources and how they value the services they receive from those
resources. For instance, coastal communities may value aesthetics very highly, particularly if they are
economically dependent on seasonal residency or coastal-based tourism, yet other coastal communities
1
Photo courtesy of USEPA
-------
focused more on fishery exploitation may value water access as an economic resource over the aesthetic
qualities of the shoreline. Thinking more broadly, the strength and type of community dependence on
natural capital may be the defining feature allowing for meaningful comparisons of what is meant by
sustainability. The critical element is a collective measure of human well-being that accounts for these
differences, but in a consistent way that allows for the examination of common ground among
communities. Such a collective measure that can suitably account for differences in how stakeholders
value available services represents a measure of how an action might contribute to human well-being
and ultimately a collective measure of sustainability.
The importance of links between of local decision making and sustainability has long been recognized in
seminal documents, including the 1987 Brundtland Commission report; the "Agenda 21" resolution
passed during the 1992 Earth Summit; and U.S. environmental policy documents, including EPA's 1999
Framework for Community-Based Environmental Protection (Portney 2001). Interest in establishing
local sustainability programs began to take root in the U.S. in the late 1980s and early 1990s in
communities like Jacksonville, Florida; Seattle, Washington; and Boston, Massachusetts (Portney 2014).
Of more than 2,000 local governments in the U.S. that responded to a 2010 International City/County
Management Association (ICMA) survey, 15.6% had a budget specifically for sustainability efforts and
26.8% had devoted staff to a sustainability efforts (ICMA 2010). Subsequent analyses of the ICMA data
found that larger municipalities were more likely to have sustainability programs than smaller
municipalities (Nye and Mulvaney 2016, Portney 2014).
Cross-sectional studies of local sustainability programs find that most communities consider the "three
pillars" of sustainability (i.e., environment, economy, and social) in strategic planning, as well as the
intersections among the pillars (i.e., livability, viability, equity, and sustainability) (Portney 2014,
Tanguay et al. 2009). A common focus of community sustainability programs is the environment,
though there is a wide range in the way that environmental sustainability is defined across local
programs. Programs emphasize environmental outcomes that the community can directly influence (e.g.,
land use), issues that are beyond a community's control (e.g., regional air quality), or issues across this
continuum. Programs also differ with respect to their focus on issues with direct local impact (e.g.,
wastewater management) to issues with global implications (e.g., greenhouse gas emissions).
A key component of more advanced local sustainability efforts is the development of "sustainability
indicators" that define the focus of the program and provide a way to measure progress. The EPA
defines sustainability indicators as "a measurable aspect of environmental, economic, or social systems
that is useful for monitoring changes in system characteristics relevant to the continuation of human and
environmental wellbeing" (Fiksel et al. 2013). In an analysis of 17 studies of local sustainability
indicators programs in the U.S., Canada, and Europe, Tanguay et al. (2009) compiled 188 uniquely
defined indicators in use by local governments. The study found the greatest consistency among
communities in the economic indicators used (e.g., employment status and income). The study found
great variation in how communities measure environmental and social sustainability. Portney (2014)
found similar patterns in a study of a national cross-section of local indicator projects.
Sustainability indicators used by local communities often serve multiple purposes and are communicated
in different ways. Sustainability indicators are used to identify and diagnose issues, raise awareness,
build grassroots support, influence local decision making, measure and communicate progress, and
manage sustainability efforts (Lubell et al. 2009, Portney 2014). Local sustainability programs may
include development of indicators, regular progress reporting based on indicators, and integration of
indicators into program evaluation through development of actionable targets (Portney 2001, Portney
2014). Progress reports can include a consistent set of indicators or can change focus over time. The
Community Indicators Consortium (CIC) maintains extensive examples of community sustainability
2
-------
indicators projects, progress reports, tools, and other resources (Community Indicators Consortium
2015). What is needed is synthesis and a consideration of how broadly suites of indicators can be
applied to the measuring success across community types.
This report focuses on four key areas for comparison of communities: community composition,
stakeholder priorities, availability and quality of ecological resources, and measures of human well-
being. First, the make-up of communities based on socio-demographic, economic, and ecological
composition are examined. This is an objective description of both resource availability and
dependencies in a community, and combines characteristics of both people and place into one
classification. Second, the community priorities as reported by the stakeholders are considered. This
comparison explores commonalities in the decision context and fundamental objectives of community
stakeholders as the starting place for decisions. Third, a geographic information system (GIS) mapping
approach is used to consider similarities and differences in the availability of important index ecosystem
goods and services (EGS). This element directly compares across communities the type and amount of
benefit humans are receiving from their environment at the local scale. Finally, similarities are explored
in a measure of human well-being as an estimate of the impact of environmental decisions on overall
quality of life. Combined, these four elements of comparison represent the major components of
decision support from factors driving decision priorities, probable pathways for environmental impacts
of decisions, and finally to probable impacts on beneficiaries. This analysis will consider each of these
four elements in turn, and then synthesize the outcome into recommendations for use of community
similarities in a national strategy for community-based decision support.
The analysis is also split into two parts based on the specific community focus. The first part is a general
examination of all coastal communities in the contiguous United States and includes the examination of
community classification and the HWBI. The second part is focused specifically on a set of four index
communities selected based on the classification system. A comparison of data from these index
communities will include stakeholder priorities and availability of EGS resources.
The objective of this comparison is to find common ground among local communities in the four
elements that may be informative regarding decision support. For the purposes of this analysis, a
community is defined as a municipality, however the influence on municipal level decisions may extend
as far out as the county level, so the data used in this analysis is confined between municipal and county
level with the specific data scale defined in each chapter. All of the data used in this analysis are applied
at county level or lower. The comparisons in this report are intended to demonstrate the value of
national-scale community comparisons for working on local-scale decision support. This report
highlights both data and approaches for this purpose in order to better support EPA goals.
3
-------
1.1 Literature cited
Abdul-Aziz, O.I. and S. Al-Amin. 2016. Climate, land use and hydrologic sensitivities of stormwater
quantity and quality in a complex coastal-urban watershed. Urban Water Journal 13:302-320.
Community Indicator Consortium. 2015. Indicator Projects, [accessed 14 September 2016], Community
Indicators Consortium.
Fiksel, J., T. Eason, and H. Fredrickson. 2013. A Framework for Sustainability Indicators at EPA. U.S.
Environmental Protection Agency, Washington, DC, EPA/600/R/12/687.
Fruet, T.K., F.G.d.S. Pinto, Y. Moretto, L.D. Weber, M.C. Scur, and A.C.d. Moura. 2016. Influence of
the land use on the water quality in the Sao Joao and Iguacu Rivers, state of Parana, Brazil:
Assessment of the importance of the riparian zone. African Journal of Agricultural Research
11:48-56.
ICMA. 2010. Local Government Sustainability Policies and Programs. International City/County
Management Association, Washington, DC.
Lubell, M., W.D. Leach, and P. A. Sabatier. 2009. Collaborative Watershed Partnerships in the Epoch of
Sustainability. In: Mazmanian, D.A. and M.E. Kraft (eds.), Toward Sustainable Communities:
Transition and Transformations in Environmental Policy, 2nd edition. MIT Press, Cambridge,
MA, pp 255-288.
Nye, M.B. and K.K. Mulvaney. 2016. Who is next? Identifying communities with the potential for
increased implementation of sustainability policies and programs. Sustainability 8.2:182.
Portney, K.E. 2001. Taking Sustainable Cities Seriously: A Comparative Analysis of Twenty-Three U.S.
Cities. 2001 Meeting of the American Political Science Association, August 30-September 2, San
Francisco, CA.
Portney, K.E. 2013. Taking Sustainable Cities Seriously, 2nd edition. MIT Press, Cambridge, MA.
Portney, K.E. 2014. Developing Sustainable Cities Indicators. In: Mazmanian, D.A. and H. Blanco (eds.),
Elgar Companion to Sustainable Cities: Strategies, Methods and Outlook. Edward El gar,
Cheltenham, UK, pp 283-301.
Tanaka, M.O., A.L. Teixeira de Souza, L.E. Moschini, and A.K. de Oliveira. 2016. Influence of
watershed land use and riparian characteristics on biological indicators of stream water quality in
southeastern Brazil. Agriculture Ecosystems & Environment 216:333-339.
Tanguay, G.A., J. Rajaonson, J.F. Lefebvre, and P. Lanoie. 2009. Measuring the Sustainability of Cities:
A Survey-Based Analysis of the Use of Local Indicators. CIRANO Scientific Publications,
Montreal, Quebec.
4
-------
2 Community classification system
The Sustainable and Healthy Communities (SHC) research program is intended to support sustainable
decision making at the community level. The SHC program defines communities as those people that
reside within the jurisdiction of one or more local governments or tribal nations; and stakeholders
include community decision makers and other groups that share interest in SHC research. Sustainability
is defined as the ability of a community to meet present needs without compromising the ability of
society and the environment to meet the economic, social, and environmental needs of future
generations. Need is defined, both present and future, as delivery of ecosystem goods and services
(EGS). The term decision is used here generally to consider all actions that may be taken by a
community that may affect the sustainability of EGS delivery. Decision support tools or approaches
encompass all of these elements and provide a link between available resources and community
objectives. Such tools, based on a common set of definitions are valuable for comparing results across
communities and community types.
Once decision tools or approaches have been developed and validated within multiple communities, the
question of their transferability and generality becomes an important element of tool utility. All
communities are different, but they may possess common elements that are informative regarding both
how a decision tool may be used and how effective a tool developed in a different community may be in
this novel application. A clear objective of transferability in decision support is to develop methods for
delineation of coastal communities that are informative regarding similarities and differences in links
between available EGS, community priorities, and the sustainability of community decisions. An
analytical community-classification system (CCS) is intended to delineate communities according to
their environmental, social, and economic composition (Harris 2010, Nye and Weden et al. 2011,
Mulvaney 2016). This CCS is a critical element for assessment of transferability and supports research
in proceeding chapters of this report through the facilitation of cross-community comparisons. In this
chapter we describe a CCS based on three distinct sets of data, which will be the basis for comparisons
among communities in all subsequent chapters.
2.1 Methods
2.1.1 Input data
The CCS was constructed from three sources of data intended to describe a community with respect to
three pillars of sustainability (social, economic, and environmental). While the CCS is not intended to
describe the sustainability of a community, it is intended to delineate communities with respect to their
priorities and available resources, and the CCS will be more informative if that description is well linked
to sustainability measures. The three data types are: social/demographic composition; employment
location quotient; and ecoregion.
The chosen measure of community social/demographic composition was the Tapestry dataset (ESRI
Corporation; accessed 27 April 2012) that is a multivariate analysis of census data at the zip code+4
level (e.g., street level; United States Postal Service; accessed 15 September 2016). Data included in this
measure include population size and density, median income, education level, age distribution, and
median home values. The raw data were transformed into summary groups with a principle components
analysis to summarize the variability into a suite of 12 orthogonal variables labeled 'dataset' (ESRI
Corporation; accessed 27 April 2012). The Tapestry data were then summarized at the county level as
5
-------
the proportion of each of the 12 Lifemode categories represented in a county. Qualitative descriptions of
the 12 tapestry categories are available from ESRI along with the dataset and are summarized in the
Results.
The employment Location Quotient (LQ) is a measure of proportion of local employment within North
American Industry Classification System (NAICS) sectors compared to the national average. The LQ is
available from the Bureau of Labor Statistics (Bureau of Labor Statistics; accessed 10 October 2012) by
economic quarter at the county level. For the purposes of this analysis the employment data were
apportioned into three categories based on NAICS supersectors. The first category was labeled 'Local
Dependence' and was comprised of employment data for forestry, fishing, agriculture, mining, oil and
gas extraction, and tourism (NAICS 11, 21, 713, and 721). Locally-dependent tourism employment was
separated from more general hospitality sectors jobs and included in the Local Dependence category, but
this was incomplete as some NAICS sectors that could not be clearly separated were excluded (e.g.,
NAICS 72 'Accommodation and food service' can be subdivided between tourism and non-tourism
components, but NAICS 48 'Transportation' cannot). The second category was labeled 'Throughput'
and represented all manufacturing (NAICS 31-33) jobs held by residents of the county. Manufacturing is
meant here to summarize employment that is only partially locally based (e.g., factory infrastructure) but
is also dependent on raw materials obtained outside the community and could be relocated and/or
replaced with another equivalent employer. The third LQ category is labeled 'Service' (NAICS 51-56,
61-62, 81) and is comprised of service sector employment not associated with tourism or the public
sector. The LQ data are comprised of three dimensionless ratio values (> 0; Local, Throughput, and
Service) and an LQ value > 1.5 is considered a deviation from the national average (Riddington Gibson
and Anderson 2006).
The final data category, 'Ecoregion', represents the environmental resources available to a community.
This is not an inventory, but an index based on a suite of environmental variables including topography,
geologic composition, and climate (U.S. Geological Survey 2012). Ecoregion data are organized into a
set of 85 categories that describe the conterminous United States and CCS input data are the proportion
of each county comprised of each category. Overall, 70 variables were used in the CCS
(Social/Demographic - 12, Employment - 3, Ecological region - 55).
The target scope of this analysis is coastal communities and is intended to be comparative and examine
features of coastal communities that are distinctive and relevant to community decisions involving
ecosystem services (Weden et al. 2011, Mikelbank 2004). This analysis is also constrained by current
availability of data. For these reasons the scope of the analysis is all coastal counties in the conterminous
United States. The scale of this analysis is the county level. This choice is partially driven by data
constraints as all information needed is available at the county level. That said, the county scale is a
useful upper limit for defining community boundaries. Many decisions at the community level are made
by county commissions (e.g., millage rates), making the county a partner in many cases. In addition,
state level dynamics are a summation of all counties within the state, so the county level of analysis can
be viewed as modular for shifting to coarser scales of analysis. The definition used here for a coastal
county is that used by NOAA (NOAA's List of Coastal Counties for the Bureau of the Census; accessed
14 September 2016) and includes 662 counties nationwide and 158 in the GOM region (Figure 2.1).
6
-------
Figure 2.1 Map indicating the counties in the conterminous United States defined as coastal counties for
this analysis (black outline). The definition is from the NOAA classification of coastal counties. Variation on
base map is Ecoregion data used in the community classification system.
2.1.2 Data analysis
The data were analyzed for a parameter reduction and delineation of county-level data into groups based
on multivariate patterns. The data from the three elements of the CCS were tabulated by coastal county
and initially analyzed with a Bayesian model-based cluster analysis to identify the most likely cluster
pattern in the dataset. The data were then analyzed with a hierarchical agglomerative cluster analysis of
Euclidean distance. The number of groups for the hierarchical analysis was derived from the Bayesian
outcome. Based on this analysis, all coastal counties were assigned to a group. A comparative analysis
of variable mean (sd) values by the group was conducted to establish general differences between the
characteristics of the groups and a mapping exercise was conducted to examine spatial distribution of
coastal counties by group. All data analysis was conducted in R with thepvclust package (R Network;
accessed 14 September 2016).
7
-------
2.2 Results
2.2.1 Cluster analysis
Bayesian model analysis indicated that the optimal grouping number for all coastal counties was eight.
Examination of the separation of groups indicates both strong (e.g., Groups 5 & 6) and weak (e.g.,
Groups 2 & 8) groups as measured by statistical distance (Figure 2.3) but clear separation. Examination
of the group loadings across the eight principal components indicated a nearly balanced influence for all
eight which supports the validity of eight cluster groups for the analysis (Figure 2.2).
PCI
PC2
PC3
PC4
PC5
PC6
PC7
PC8
SS LOADINGS
3.213
2.495
2.412
2.175
2.105
1.956
1.774
1.590
PROPORTION
0.058
0.045
0.044
0.040
0.038
0.03
0.036
0.029
VARIABLES
CUMULATIVE
0.058
0.104
0.148
0.187
0.225
0.261
0.293
0.322
VARIABLES
6 n
4 -
Locale
Service
Throughput
Principal component 1
Figure 2.2 Scree plot indicating the distribution of 663 coastal counties with respect to the first two
principal components. Red arrows indicate association and strength of variables for each principal component.
The Lifemode variables are indicated by LI through L12, the LQ variables are labeled Local, Throughput, and
Service, and the Ecoregion variables are included in the analysis but not shown here for clarity. The inset table
displays the loadings for all eight of the principal components along with the proportion of variance explained by
each one. Lifemode labels are: High society (LI), Upscale avenues (L2), Metropolis (L3), Solo acts (L4), Senior
styles (L5), Scholars and patriots (L6), High hopes (L7), Global roots (L8), Family portrait (L9), Traditional living
(L10), Factories and farms (LI 1), and American quilt (L12). Lifemode descriptions can be found in the Tapestry
segmentation reference guide (G53769; ESRI; accessed 19 September 2016).
8
-------
CD
O
+
CD
xr
CD
o
+
CD
CO
CD
O
+
-------
The relationship between social and economic variables was particularly evident in the analysis.
Examination of the scree plot indicated the three economic dependence measures (Throughput, Local,
and Service) were nearly orthogonal to each other. High local importance of service-based jobs loaded
heavily with the urban, affluent, and ethnically-diverse Lifemodes. High dependence on local resources
loaded closely with rural and young family Lifemodes, and while Throughput loaded well with rural
communities as well, this economic variable was more closely associated with high median age and a
salaried workforce (Figure 2.2). No clear trend was evident for environmental data with respect to the
other two data elements as the Ecoregion variables were equally distributed across all principal
components (Figure 2.2). There were both economic and demographic trends among the groups, but
environmental variables were generally well balanced reducing their influence on group membership.
The number and distribution of counties by group differed wi dely among groups (Table 2.1). In
particular, Group 4 includes only three counties and Group 7 is almost completely within the state of
Florida. Social variables such as population size and median home value showed a clear trend among
groups with Groups 1 and 3 including all of the counties with a population density over 5,000 mi"2
(Figure 2.4a). The remaining groups displayed minimal variability with respect to population size
indicating the most urban areas are all in Groups 1 and 3. The Economic variables also showed group
bias with Group 4 having the highest value for Local Dependence (21.4) followed by Groups 6(11.5)
and 5 (6.8). All other groups had LQ values near or below the 1.5 threshold. The Throughput LQ score
only exceeded 1.5 for Groups 8 (3.6) and 2 (1.8), and no group had a Service LQ score greater than 1.2
(Figure 2.4b; Table 2.1). Environmental variables were very balanced and not very influential on group
membership with the exception of Group 7, which is heavily weighted towards Ecoregions in central
Florida and the south Atlantic coast (Table 2.1; Figure 2.5). There was however, an evident geographic
distribution of groups with Group 1 dominant in the northeast, Group 2 in the Midwest including the
Great Lakes region, and Group 3 in Texas and California.
Photo courtesy of U SEP A
10
-------
60000
~ 50000
40000
30000
c
0)
T3
C
o
™ 20000
a.
2 10000
0
New York City
i :
1—I—I—~—~—~—#—~—i
o
8
Figure 2.4a Summary plots showing the distribution of mean population density (mi-2) for coastal counties
by Group (A). Outlier County labeled for clarity.
OJ
o
u
l/>
4->
c
Q)
5
o
3
cr
c
o
5
to
u
o
25.00
20.00
15.00
10.00
5.00
0.00
i
n1 i1* lii
1:
n ii1 i -
B
Local
Throughput
Service
12345678
Typology Group
Figure 2.4b Summary plots showing the mean (SD) LQ scores for all three categories by Group (B).
11
-------
Group 1
Group 2
Group 4
Group 5
Group 6
Group 7
Group 8
Figure 2.5 Map of coastal counties with group membership indicated by color. Legend gives
group number and a qualitative description by group number is available in Table 2.1.
12
-------
Table 2.1 Summary of community classification system data and results by group including a qualitative description of each group based on the
data (Description), number of counties included in each group (n), states with counties in each group (flagged if only one county), highest and
lowest Lifemode (LM) category for each county by percentage, mean (SD) LQ scores for each county by employment category (Throughput, Local
Dependence, and Service), and the maximum percentage among the 85 Ecoregion categories indicating the evenness of coverage across the eight
groups. The LQ scores in bold indicate values greater than 1.5 the national value.
GRP
Description
n
States
Highest LM
(%)
Lowest LM
(%)
Throughput
Local
Dependence
Service
Ecoregion
Maximum
(%)
1
Largest populations
and diverse in LM's
223
AL, CA, CT, DE, FL, IL, IN, LA, ME, MD,
MA, Ml, MS, NH, NJ, NY, NC, OH, OR,
PA, SC, TX, VA, WA
High society
Working
class, small
comm.
0.8 (0.3)
1.1(0.4)
1.2 (0.3)
13.6
2
Working class, rural,
manufacturing, older
129
AL, CA, DE, GA, LA, MD, Ml, MN, MS,
NY, NC, OH, PA, SC, TX, VA, WA (1), Wl
Middle age,
middle
income
Ethnic
diversity
1.9 (0.6)
1.4 (0.6)
1.0 (0.2)
24.9
3
Suburban with high
Local Dependence
133
AL (1), CA, DC, FL, GA, LA, ME, MD,
MA, Ml, MN, MS, NH, NJ, NY, NC, OH
(1), OR, PA, SC, TX, VA, WA
n/a
n/a
0.9 (1.0)
3.1 (0.6)
1.1 (0.4)
13.7
4
Ethnic, young, max
Local Dependence
3
CA (1), FL
Ethnic
diversity
Young
families
0.4 (2.0)
21.4 (0.03)
0.8 (0.4)
50.2
5
Working class, high
local dep., younger
families
45
CA, FL, GA, LA, Ml, MS (1), NC, OR (1),
TX, VA, WA (1), Wl (1)
n/a
n/a
1.1 (1.2)
6.8 (0.7)
0.9 (0.3)
21.3
6
Most rural, lower
income, young
families, high Local
Dependence
15
CA (1), FL, Ml, NC, TX, VA(1), WA (1)
Young
families
Urban
singles
0.9 (1.6)
11.5 (0.8)
1.0 (0.5)
23.4
7
Senior, upscale
suburban
46
FL, GA, LA (1), MS(1), SC (2)
Senior
lifestyles
n/a
0.6 (0.6)
1.5 (0.4)
1.0 (0.3)
94.5
8
Working class, small
town, manufacturing-
local
25
GA, IN, LA, Ml, MS (1), NC, OH, SC,
TX(1), Wl
Working
class, small
comm.
Senior
lifestyles
3.6(1.1)
2.0 (0.7)
0.7 (0.2)
21.7
13
-------
2.2.2 Group description
Based on the analysis of key scores, groups were given a qualitative description to aid in interpretation
of group membership (Table 2.1). Groups 1, 2, and 3 were the largest in terms of number of counties
(Table 2.1). Groups 1 and 3 also contained the largest variability in population density (Figure 2.4a).
Group 1 had the highest score for affluence and urbanization and lowest for rural community types.
Group 3 was highest for suburban/small town categories and higher age groups including retirees. Both
Groups 1 and 3 had LQ scores near 1.0 for Service and Throughput indicating they are well in line with
the national averages. Group 3 had an LQ score of 3.1 for the Local Dependence category suggesting a
higher dependence on natural resources in these counties. Group 2 scored high as working class, small
town suggesting a lower mean income, lower median education, and a lower population density than
Groups 1 and 3. The LQ scores for Group 2 were average for Local Dependence and Service, but higher
(1.9) for Throughput suggesting the increased importance of manufacturing in the Group 2 counties.
Only Groups 2 and 8 had Throughput LQ scores above 1.5 and Group 8 also included a high LQ score
for Local Dependence and included fewer counties. Groups 3-6 all had a Local Dependence LQ score
above 3.0 and showed an increasing trend in the Local Dependence LQ score suggesting a gradient for
this characteristic across these groups (Table 2.1). Group 3 was the lowest of this set at 3.6, but
increased in order for Groups 5, 6, and 4. Groups 5 and 6 displayed both high Local Dependence, as
well as a high proportion of working class and rural characteristics. These two groups only differed by
degrees of Local Dependence and 'ruralness' and were otherwise very similar. Group 4 had the highest
level of Local Dependence and also the highest level of ethnic diversity reflecting the importance of
immigration and farming in these counties. This was also the smallest group with only three counties.
However, the LQ score for Local Dependence was greater than 21, indicating a substantially higher
dependence on natural resources in these counties that accounts for the large 'distance' between Group 4
and other Groups (Figure 2.2). Finally, Group 7 was the most geographically distinct with most of the
counties in central or eastern Florida (Figure 2.5). Group 7 counties were high in affluence, but also high
in rural categories reflecting a lower mean population density and variability than observed in Group 1
(Figure 2.4a). Group 7 also included the highest mean score in percentage of senior residents and a
higher median age in these counties. Group 7 counties also scored near the bottom in ethnic diversity
and young families with children. The LQ scores were all < 1.5 suggesting average employment patterns
overall in Group 7.
2.3 Discussion
The CCS developed for this study delineates coastal counties well both by region of the country and
within a region. The primary delineation is associated with population density in that highly-dense
counties all fell into Group 1. As population density drops, the county type begins to become more
specialized and as a result becomes easier to delineate. After population density, the next most
informative level of delineation was primarily a function of employment LQ scores. The LQ scores
indicate how specialized the local employment patterns are by comparison to the national average.
Counties that display a high LQ score in one of the three LQ categories (Local, Throughput, and
Service) are highly dependent on that category, at least for number of jobs. Counties with a low
population density displayed an increasing trend in the Local Dependence LQ score suggesting a trend
in dependence on local natural resources such as farming, fishing, mining, or oil and gas extraction.
Tourism is also included in this LQ category, but its importance is hard to gauge as increased tourism
employment is thought to go hand in hand with service sector jobs so should affect both Local
Dependence and Service categories. As Local Dependence increases the most likely contributors are
agriculture and fishing jobs. Local Dependence as much as 21 times the national average was observed
within single counties, but most of these locally dependent counties were spread out with the largest
14
-------
cluster in Louisiana and eastern Texas. In point of fact, Louisiana was the most diverse state with both
high Local Dependence, high Throughput, and close to the full range of population densities.
Overall the demographic makeup of the counties was in line with the employment and population size
patterns. More urban and suburban counties by population density were also more diverse in terms of
ethnicity, affluence, and family composition. As communities became more dependent on certain
employment sectors (higher LQ scores) and less dense they also became more demographically
specialized suggesting a shift that comes along with higher Local Dependence. This finding has not
been thoroughly evaluated but will be an important element of future study.
The effectiveness of the CCS for delineating coastal counties is constrained by the choice of variables to
include. The intent of the variables included in the CCS was a link to both community priorities and
available ecosystem services. The link to natural resources may be the weakest component of the CCS
as Ecoregion was least important variable in the delineation at the county level as indicated by the
loadings in the PCA. This may be because the Ecoregion variables differ more at larger scales and are
not very important at the county level particularly along the coast (Figure 2.1). Inclusion of more 'non-
coastal' counties may change this finding. The link to ecosystem goods and services is associated with a
community's dependence on local resources and should influence a community's sense of importance
for EGS. This has not been established and one major revision to the CCS will likely involve
incorporation of a more informative measure of available natural resources, such as that described in
Chapter 5. Another limitation of the current analysis is the use of data at the county scale in the CCS.
Counties are typically diverse including both high and low density areas. As a result, the 'type' that best
describes a county is an average and the result can be highly bifurcated in extreme cases between urban
and rural areas (e.g., Palm Beach county contains both the richest and poorest people in Florida). This is
the most likely cause of the negative trend between specialization and population density. However, the
county scale is highly relevant to community decision making even for local municipalities. Analysis of
specific communities will require a 'zooming in' of the data from the county scale, but this represents a
viable mid-point between the scale of a communities decisions and the scale of their effects. Future
work will also look critically at scale and whether the CCS should be applied across scales to better
capture change.
The delineation based on the CCS is only partially consistent with the index of Human Well-Being
(Smith et al. 2013). Overall, the mean HWBI score differed by only 5-6% across the CCS groups
suggesting a weak trend at best. The variability in the HWBI nationwide is small however (< 10%), so
small changes can reflect important differences in well-being. There was some pattern to the differences,
small though they may be, suggesting that as Local Dependence increases the HWBI elements go down.
Of the three HWBI elements, the environmental element showed the most variability across groups, but
this was mainly the ranking of Groups 6 and 7, which were higher and lower respectively compared to
the overall score and the other two element scores. The meaning of this change is unclear and represents
an area for future study.
As the broader analysis of transferability of tools between communities moves forward, the next steps in
the CCS analysis will include verification of county descriptions, as well as extended comparisons of
this coastal CCS to other similar national-scale analyses and metrics. In the next chapter we will
specifically examine how a measure of human well-being differs among CCS groups. Overall, the CCS
is a useful tool for delineating coastal counties with respect to the three tiers of sustainability,
particularly along an urban to rural gradient as indicated in Figure 2.4. The limitations of working at the
county level need to be addressed and additional analysis needs to be expanded to include non-coastal
15
-------
communities. The level of information contained in the CCS is currently undetermined, but it shows
promise as a useful tool for the purpose of gauging the transferability of decision tools among coastal
communities of similar type.
2.4 Literature cited
Cutter, S.L., L. Barnes, M. Berry, C. Burton, E. Evans, E. Tate, and J. Webb. 2008. A place-based model
for understanding community resilience to natural disasters. Global Environmental Change-
Human and Policy Dimensions 18:598-606.
ESRI Corporation. 2012. Tapestry Segmentation Dataset. [accessed 14 September 2016], ESRI.
Harris, R. 2010. Meaningful types in a world of suburbs. In: Clapson, M. and R. Hutchison (eds.).
Suburbanization in Global Society 10:15-47.
Khazai, B., M. Merz, C. Schulz, and D. Borst. 2013. An integrated indicator framework for spatial
assessment of industrial and social vulnerability to indirect disaster losses. Natural Hazards
67:145-167.
Mikelbank, B.A. 2004. A typology of US suburban places. Housing Policy Debate 15(4):935-964.
National Oceanic and Atmospheric Administration. 2012. NOAA's List of Coastal Counties for the
Bureau of the Census Statistical Abstract Series.
Nye, M. B. and K. K. Mulvaney. 2016. Who is next? Identifying communities with the potential for
increased implementation of sustainability policies and programs. Sustainability 8.2:182.
Riddington, G., H. Gibson, and J. Anderson. 2006. Comparison of gravity model, survey and location
quotient-based local area tables and multipliers. Regional Studies 40.9:1069-81.
Smith, L.M., J.L. Case, H.M. Smith, L.C. Harwell, and J.K. Summers. 2013. Relating ecosystem services
to domains of human well-being: Foundation for a US index. Ecological Indicators 28:79-90.
U.S. Bureau of Labor Statistics. 2012. Location Quotient Calculator, [accessed 17 October 2016],
Location Quotient Calculator.
U.S. Geological Survey. 2012. Bailey's Ecoregions and Subregions of the United States, Puerto Rico, and
the U.S. Virgin Islands.
Weden, M.M., C.E. Bird, J.J. Escarce, and N. Lurie. 2011. Neighborhood archetypes for population
health research: Is there no place like home? Health and Place 17(l):289-299.
-------
3 Human well-being across community
types
Human activity has growing impacts on the natural capital humans depend upon for existence (Condie et
al. 2012, Pauly et al. 1998, Peterson et al. 2003). Many of these impacts are regional, national or
international in scope such as air pollution (Likens et al. 1996) and climate change (Nelson et al. 2013,
Piazza et al. 2010). Yet, there is an increasing understanding that decisions made at the local community
level can have significant impacts and need to be understood both for their local impacts, as well as for
their cumulative impacts across multiple communities (Israel et al. 1998, Tallis et al. 2008). Natural
capital degradation because of human activity is more often being valued and measured in terms of its
direct impact on human beneficiaries based on the production and supply of ecosystem goods and
services (EGS) (Garcia-Llorente et al. 2011, Grabowski et al. 2012, O'Higgins et al. 2010). All
communities have unique characteristics, but also have characteristics in common, such as beneficiaries
(i.e., resource user groups), and can be classified into groups to aid in prioritizing conservation and
utilization of natural capital. The community classification system (CCS) developed in this report
(Chapter 2) is informative about a community's priorities, and the association of these priorities with
human well-being provides a potentially valuable tool for measuring success in achieving local
objectives and informing decision makers about sustainable decision outcomes in a community-specific
context. This approach can aid decision makers in defining meaningful change in human benefit across
different communities by establishing reference points and can provide a clear justification for investing
in conservation, mitigation, and restoration of natural capital (Adeel and Safriel 2008, Pascual et al.
2012, Vaissiere et al. 2013).
Describing environmental degradation in terms of human endpoints also fosters discussion on tradeoffs
and the concept of ecosystem sustainability. Loss of natural capital has differing values for different user
groups; managers must rectify these conflicts in the context of other forms of capital (i.e., built,
economic, and social) into a coherent plan that considers the synergistic outcome for all user groups
(Butler et al. 2013, Green et al. 2014). Such a plan must also include measurable reference points to
evaluate changes in capital as meaningful to beneficiaries. The most useful end point for this approach is
the concept of ecosystem sustainability, which rather than focusing on each beneficiary individually,
targets the maintenance of net benefits through time (Jorgenson et al. 2014). This is still an 'ecosystem-
centric' approach; nonetheless it is dependent on understanding dependencies of human benefits to a
broad range of EGS, as well as defining a clear and acceptable measure of overall sustainability
(Abunge et al. 2013, Yang et al. 2013). Classifying communities in terms of their economic and social
dependence on current delivery of EGS yields a potentially informative way of delineating communities
for the purpose of establishing local reference points from which to measure change, if a community's
type can be linked to differences in community sustainability.
Measuring sustainability requires knowing what stakeholders wish to sustain. Net delivery of EGS to
humans provides a working model for sustainability but currently lacks a coherent framework.
Numerous measures of sustainability exist (Krotscheck et al. 2000, Putzhuber and Hasenauer 2010), but
most are single issue indicators, not necessarily tied to multiple human beneficiaries (e.g., Neset and
Cordell 2012, Velasquez et al. 2011). Suites of indicators used to holistically measure human well-being
(HWB) show promise as a synergistic measure of the outcome of net EGS production and delivery to
humans (Smith et al. 2013b, Summers et al. 2012). Indices of HWB are a measure of benefit to humans,
beyond just economic benefits, that are also more responsive to changes in EGS production (Canadian
Index of Wellbeing 2012, Smith et al. 2013b). Indices of HWB include metrics of social cohesion, living
17
-------
standards, personal safety, civic engagement, and connections to nature (Smith et al. 2013b and cites
therein). Yet, HWB indicators are not an easily understood concept, and are not a direct measure of
service delivery. The challenge in applying HWB measures at the community level is in linking such a
broad indicator to community-specific issues and values. Different communities have different social,
economic, and environmental dependencies, defined here as the three pillars of sustainability (NRC
2011), and communities require a demonstration of HWB I utility for measuring outcomes and a
connection of HWB measures to local conditions. It is from these local index reference points that
meaningful change in HWB can be measured. Commonalities do exist in the priorities and resources of
communities and an examination of composite measures, such as a human well-being index (HWBI)
(Smith et al. 2013a, Smith et al. 2013b, Smith et al. 2012) across community types, is an effective
method to connect human well-being to community decision making.
In this chapter, the community classification system (CCS) is evaluated by combining it with the HWBI
to find well-being references points. The objective is to ask whether the CCS is informative regarding
HWBI reference points by examining whether HWBI-type indicator values differ by community type as
a potential measure of sustainability. This comparison involves the 664 coastal counties described in
Chapter 2. Additionally, the value of locally obtained data on community priorities is evaluated for the
calculation of the HWBI by integrating data from the community workshops (described in Chapter 4).
This more specific comparison will examine how local priorities may be used to alter the weight given
to the different domains of the HWBI. The overall goal is to identify associations between local
social/economic dependence on EGS and differences in human well-being that may suggest informative
local reference points for decision making about EGS provisioning. The expectation is that community
types will differ in their well-being and these differences will provide local well-being reference points
informative for measuring changes in well-being. The outcome will be an understanding of how
community classification based on EGS can be used to inform community decision making focused on
sustaining or improving HWB.
3.1 Methods
3.1.1 Classifying communities
The community classification system used in this analysis is described in Chapter 2. It was constructed
from three sources of data intended to describe a community with respect to three pillars of
sustainability (social, economic, and environmental). While the CCS is not intended to describe the
sustainability of a community, it is intended to delineate and describe communities with respect to their
priorities, dependencies, and available natural resources.
3.1.2 Measuring human well-being
The Human Well-Being Index (HWBI) used in this study is a previously described U.S. index and is a
composite of multiple indicators that characterize eight domains of human well-being: Connection to
Nature, Cultural Fulfillment, Health, Education, Leisure Time, Living Standards, Safety and Security,
and Social Cohesion as described by Smith et al. (2013b and 2012) and briefly summarized here. Each
domain is described by indicators representing a combination of metric values. For each HWBI metric,
objective and subjective data collected from various publically accessible sources were organized
hierarchically by spatial and temporal resolution (e.g., national, regional, state, and county by year) for
the years 2000-2010. When multiple spatial scales existed for a metric, the finest scale (e.g., county
versus state) was selected for processing. Data source determination was primarily driven by temporal
and spatial coverage, data reliability and credibility, historic data continuity, and future data
accessibility. All data were standardized on a scale from 0.1 to 0.9 following the Organization for
18
-------
Economic Co-operation and Development's (OECD) Better Life Index (OECD Better Life Index;
accessed 14 September 2016) approach. A detailed description of the metrics used in the calculation of
the HWBI can be found in the report entitled "Indicators and Methods for Constructing a U.S. Human
Well-Being Index (HWBI) for Ecosystem Services Research" (Smith et al. 2012). The HWBI values
were derived from indicator scores calculated as the population weighted average of the standardized
metric values. Indicator scores were averaged to create each domain score. Finally, a scaled geometric
mean was calculated across domain scores to produce the final inputs for the HWBI. Higher HWBI
scores indicate greater levels of well-being. Methods have also been developed to incorporate
community priorities into the index calculation by applying relative importance values (RIVs) as domain
weighting factors (Smith et al. 2013a). The county-level HWBI values used in the analysis of all coastal
counties were unweighted values.
3.1.3 Stakeholder-derived weightings for domains of human well-being index
Data were collected on stakeholder-derived weightings of the eight domains of the HWBI during a
series of community workshops held in nine communities across the U.S. The workshops were designed
to meet the following data quality objectives: validity, reliability, representativeness, and completeness.
Critical to all of these objectives were efforts to ensure that a broad range of community voices were
included in a meaningful way in the discussion of community values. The following subsections
summarize the workshop design and implementation, as well as participation outcomes.
Overview of workshop design
The workshops were designed to produce reliable information about community priorities. The
workshop design was developed based on the structured decision making approach (Gregory and
Keeney 2002, Carriger et al. 2013), modified to account for the limited time available for each
workshop. Part of each of the workshop was focused on exploring the relative importance of the eight
domains of the HWBI using a series of structured discussions and exercises. A list of goals reflecting the
domains of the HWBI (Smith et al. 2012) was used to frame these activities and enable comparison
across workshops.
This structured approach was intentionally designed to introduce the participants to categories of
community priorities in a stepwise fashion using non-technical language applied in a familiar context.
Another critical element of the design of the workshop was the selection of participants to best represent
the community. This affects the extent to which the HWBI data gathered in the workshops accurately
represents the priorities of the community as a whole. The project team sought to maximize
representativeness of workshop participants in the planning and implementation of the workshops.
Workshop implementation
Participants were provided with the list of domains used in the HWBI (Smith et al. 2013b) and were
given time to review and ask clarifying questions about the domains. In small groups, participants were
asked to map the qualities identified in the previous discussion to specific HWBI domains.
Using a multi-voting (dot voting) process, each participant was given the opportunity to identify the
goals most important to him or her. Facilitators placed flip charts with the lists of goals on a wall in the
room and participants were each given seven dots. Participants were asked to place their dots next to the
goals that they felt were most important to the community. Participants were allowed to distribute the
seven dots next to one or more goals, placing as many dots next to a goal as desired.
19
-------
As a final exercise for the first part of the workshop, participants were asked to rank goal categories
(e.g., Health, Education) based on their individual views of how important each category is to the well-
being of members of the community. Participants were asked to use a scale from 1 (most important) to 8
(least important) with no ties.
The priority data collected during these workshops was used to consider changes to the RIVs used in
calculation of the HWBI. A comparison of weighted and unweighted HWBI values in workshop
communities was conducted based on data obtained during community workshops. These data were
used to calculate RIVs for all eight domains and the HWBI was recalculated based on these RIVs for the
nine counties in which a workshop was held. The recalculation of the HWBI was the weighted average
of raw domain scores based on the workshop-based RIVs.
The unweighted HWBI was compared among CCS groups to examine patterns in well-being as a
function of the community classification. The comparison was conducted for the overall HWBI and also
separately for the eight domains of the HWBI. Patterns in the outcome were then examined with respect
to CCS group characteristics. All statistical comparisons were conducted with a 1-way analysis of
variance (ANOVA) with a type-I error rate of 5%. If the ANOVA results were significant, this test was
followed by a Turkeys HSD multiple comparison to test for between group differences while preserving
the experiment-wise type-I error rate (Zar 2010). The weighted HWBI was calculated from stakeholder
engagement results for individual communities and these locally-weighted HWBI values were visually
compared to the unweighted national HWBI.
3.2 Results
3.2.1 Human well-being and community type
Human well-being index values showed significant differences among CCS groups (ANOVA F7,6ii=
6.6, p< 0.001). Post hoc comparisons indicated that the mean HWBI for Group 1 was significantly
higher than Groups 4, 6, and 7 (Figure 3.1). Analysis of the eight domains of the HWBI (Figure 3.1)
identified more specific differences between the CCS groups with respect to well-being elements. No
significant difference was observed between classification groups for the Connection to Nature or
Cultural Fulfillment domains. The remaining six domains indicated a significant difference at a= 0.05
with five significant at the Bonferroni adjusted a= 0.0063. For the Health domain, Groups 1, 2, and 3
were significantly higher than Groups 7 and 4 (Tukeys HSD, adjusted p< 0.05) and Groups 8 and 5 were
also significantly higher than Group 4 (adjusted p< 0.05). Group 4 had the lowest score for Health. For
the Leisure Time domain, Group 4 had the highest score and was significantly higher than Groups 6, 8,
2, and 5. Groups 7 and 1 also had a significantly higher Leisure Time score than Groups 6 and 8. Groups
5 and 2 were significantly higher than Group 6 for the Leisure Time domain. For the domain Living
Standards, Group 1 had a significantly higher score than all other groups except Group 3. For the
domain Safety and Security, Group 1 had a significantly higher score than both Groups 6 and 4. For the
domain Social Cohesion, Group 6 had a significantly higher score than Groups 4, 7, 5, and 1 (Figure
3.1).
20
-------
65 1 A
60
o
o
c/)
c 55 -
(D
.O
1
c
03
JL
50 -
45 -
HWBI
Connection to Nature
Cultural Fulfillment
Education
Health
Leisure Time
Living Standards
iinimi Safety and Security
D Social Cohesion
~~T~
4
i
5
i
6
1
Classification group
Figure 3.1 Summary of the human well-being index (HWBI) scores by community classification group. See
Chapter 2 for details on the community classification system categories. Mean (sd) values are presented for the
overall index (A) and z scores are presented separately for the eight domains of the HWBI (B), which describe
deviation from the overall mean value.
21
-------
3.2.2 Weighted vs. unweighted human well-being index
Weightings of HWBI domains based on stakeholder workshop data showed some clear trends across all
communities (n= 9). The Education, Health, and Social Cohesion domains were consistently the top
choices across all communities (Table 3 .1). In contrast, Connection to Nature and Cultural Fulfillment
were consistently the lowest ranked domains although Windsor Locks ranked Connection to Nature
higher than the other eight communities.
The weighted HWBI score was higher in four of the nine communities examined (Table 3.2; Figure 3.2).
The maximum increase occurred in Harrison County, IA (+3.2%), where the top two ranked domains
(,Education and Health) were also in the top three raw HWBI scores. In contrast, the weighted HWBI
score had the largest drop compared to unweighted in St. Landry Parish, LA (-10%). In this community,
over twice as much weight was placed on Education as on the other seven domains and Education had
the lowest raw HWBI score. This disconnect resulted in a large drop in the overall weighted HWBI.
Only one other community saw a change in the weighted HWBI of more than two percent (Forsyth
County, NC; -2.2%) and again here the Education domain was ranked highly but had the lowest raw
score.
Photo courtesy of U SEP A
22
-------
Table 3.1 Relative frequency of goal categories receiving dot votes during workshop group voting exercise. Goal categories are identical to the
domains of the human well-being index (HWBI).
HWBI
Domain
Woodbine,
IA
Pascagoula,
MS
Lewisville,
NC
Pensacola,
FL
Vero
Beach,
FL
Freeport,
NY
Windsor Locks,
a
Thibodaux,
LA
Opelousas,
LA
Education
24
5
14
24
16
26
14
22
17
Health
15
16
16
14
21
9
20
12
18
Work life
balance
11
7
13
4
15
7
12
13
14
Living
standards
15
11
11
21
13
15
5
14
15
Safety and
security
10
12
16
13
8
17
8
14
12
Connection
to nature
8
5
9
2
7
5
13
2
5
Cultural
fulfillment
2
11
4
4
7
5
7
3
4
Social
cohesion
16
21
19
19
14
16
18
21
14
23
-------
Table 3.2 Summary of weighted (Wt) and unweighted (Uwt) scores for the human well-being index (HWBI). Labels in the first column are the eight
domains of HWBI. See text for details.
Community
Pensacola
Escambia
Co FL
Vero
Beach/Indian
River FL
Thibodaux/
Lafourche
Parish LA
Opelousas/
St. Landry
Parish LA
Pascagoula/
Jackson Co.
MS
Woodbine/
Harrison Co. IA
Lewisville/
Forsyth Co.
NC
Freeport/
Nassau Co.
NY
Windsor
Locks/
Hartford
Co, a
Domain
Wt
Uwt
Wt
Uwt
Wt
Uwt
Wt
Uwt
Wt
Uwt
Wt
Uwt
Wt
Uwt
Wt
Uwt
Wt
Uwt
Connection
3.4
6.1
5.4
7.2
4.1
7.2
6.1
10.6
3.7
6.1
2.5
4.3
4.3
7.2
4.4
6.7
6.7
6.7
to nature
Cultural
4.0
6.4
4.0
5.1
3.8
6.2
3.9
6.0
5.8
6.5
4.8
7.6
4.6
5.1
5.1
7.3
5.5
6.7
fulfillment
Education
8.7
5.8
4.8
4.7
6.6
4.2
8.5
3.6
7.0
6.3
13.9
7.5
4.8
4.3
7.3
5.5
4.9
5.3
Health
8.2
7.2
8.3
7.3
6.2
7.6
8.0
7.1
7.2
7.1
10.6
7.6
7.8
7.7
7.0
8.0
10.9
7.7
Leisure
4.6
7.5
6.0
7.1
5.1
6.6
5.7
6.9
4.4
6.9
4.9
6.7
4.9
7.7
4.3
6.9
5.2
6.9
time
Living
6.3
6.2
6.8
6.6
7.0
6.6
5.0
6.0
8.0
6.3
5.6
6.5
7.7
6.1
7.5
7.5
4.7
6.8
standards
Safety and
7.5
6.2
9.6
7.2
10.3
7.4
4.5
5.7
8.8
7.0
8.7
8.4
8.7
6.9
14.5
9.5
9.4
8.0
security
Social
7.2
5.4
5.8
5.2
6.3
5.1
4.0
4.8
7.1
5.9
6.4
6.9
6.1
5.0
6.9
5.3
6.4
5.2
cohesion
HWBI
49.8
50.7
50.7
50.4
49.4
51.0
45.7
50.7
52.0
52.0
57.4
55.6
48.9
50.0
56.9
56.7
53.6
53.3
overall
24
-------
60
59
58
Thibodaux Vero Opelousas Pensacola Freeport Pascagoula Woodbine Lewisville Windsor
Community names
¦ Priority Wted HWBI ~ Unwted HWBI
Figure 3.2 Comparison of mean weighted (wted) and unweighted (unwted) values of the human well-being
index (HWBI) among the nine communities included in the study. Communities in the study are described in
the text.
3.3 Discussion
Human well-being is a diverse topic that is increasingly utilized in assessing community sustainability
(Jorgenson et al. 2014, Smith et al. 2013b), and is a viable alternative to monetary valuation for
informing decision makers about the outcomes of their decisions in terms of worth to human
beneficiaries (Tallis et al. 2008). However, it lacks a consistent definition and a clear reference point
from which communities can measure change. Here the concept of HWB is linked to community type to
inform the process of establishing reference values for HWB and also to better understand how a
community's static features (e.g., composition and resources) impact HWB indicators. First, an index of
community type is described and then one measure of HWB is examined as a function of community
type.
The intended use of the HWBI is to evaluate the influence of social, economic and ecological service
flows on human well-being and consequently how human well-being may change between communities
and through time (Summers et al. 2012). Geographic differences in county-level HWBI values similar to
the ones described here have been reported elsewhere (Smith et al. 2013b) and indicate these values may
include important differences associated with different types of communities. Understanding these
differences yields community-specific reference points for measuring meaningful change. High well-
being scores were associated with high population density and with low dependence on local resources.
The groups with the highest Local Dependence scores also had the lowest overall well-being score. In
contrast, high economic dependence on manufacturing was not associated with lower overall well-being.
The importance of population density may be an artifact of diversity as it has been reported elsewhere
that population centers foster higher overall well-being because of access to resources (Smith et al.
2012). One surprising outcome was that this indicator of well-being did not associate well with
community types high in individual affluence. The CCS group with the highest proportion of affluent
25
-------
citizens was Group 1, which had a high well-being score, but the group with the next highest affluence
score ranked near the bottom in well-being; and Group 8, with a well-being score nearly as high as
Group 1, had the second lowest proportion of affluence. This is, of course, a cumulative measure of
well-being that contains environmental, economic, and social indicators, and a cumulative measure
should be most responsive to socio-economic diversity. This is an interesting pattern in the HWBI across
community type, and separation of the cumulative score into its requisite categories may be informative
about how available resources are associated with well-being, particularly for the smaller more
homogenous CCS groups.
The overall HWBI score for a county can be separated into eight domains (Table 3.1) all of which
produce separate scores that are combined into the composite score. Of the eight, five were
meaningfully different among CCS groups with the most important separations between the high
population density Group 1 and the rural, high Local Dependence Groups (4 and 6). The ranking of
community types differed by well-being domain. The most urban group had the highest score for all
elements of well-being except Leisure Time, Connection to Nature, and Social Cohesion. These three
were all positively associated with decreases in population density and increases in resource
dependence. The local resource dependent groups were split for the domain Leisure Time with the
higher LQ Group 4 having the highest score for this domain and Group 6 the lowest. Group 4 is really a
sub-group of 6 with the key differences associated with ethnic diversity and there may be a cultural
factor in this separation, however this is not testable here. Leisure Time scores were above average for
the high population density Group 1 and the high median age Group 7. This pattern is consistent with
expectations and supports the association of more Leisure Time in areas with a high proportion of
affluence, but the Leisure Time score was still higher in the more rural, resource dependent counties.
The domains for Living Standards and Safety and Security were both heavily weighted to population
density and diversity with clear separation between the urban group and the rest of the CCS groups.
Nonetheless, the Safety and Security domain was another clear delineation of Group 1 between the rural,
high Local Dependence groups. Safety and Security was significantly lower in Groups 4 and 6,
suggesting an association of social vulnerability with rural resource dependent communities. This seems
counterintuitive on the surface as low Safety and Security is often associated with urban environments
(Gilbert 1999), but this domain is described by number of accidents, impacts of hazardous weather, and
social vulnerability in combination with crime rates per 100,000 people. The results suggest the impact
of the non-crime categories on Safety and Security cannot be dismissed with respect to well-being.
Social Cohesion was expected to be highest in rural, specialized communities, and Group 6 had a
significantly higher score for Social Cohesion than the denser and less locally dependent CCS groups.
Social Cohesion was also high in the other two rural Groups (2 and 8) and low in high density Group 1.
However, the smaller Group 4 had the lowest score for Social Cohesion of all CCS groups. Again this
may be related to the combination of high local resource dependence and higher than average ethnic
diversity. Social Cohesion is based on expressions of trust, political engagement, membership in
community organizations, and volunteerism. The counties in Group 4 are all in California or Florida and
strongly tied to agriculture, so the higher ethnic diversity may suggest a migrant labor force, which may
reduce Social Cohesion, at least as it was measured in this study. Again this is a clear indication that
although this group is small, the combined differences detected between groups with high resource
dependence are real and important for delineating human well-being reference points for measuring the
local impacts of decisions on HWB.
Overall, U.S. coastal counties increase in well-being as they increase in population density and socio-
economic diversity. Yet, the more rural counties along the coast had higher scores for specific domains
of well-being, such as Social Cohesion and Leisure Time suggesting there are elements of well-being not
associated with high population density. However, urban centers had well-being scores that were more
26
-------
consistently above average across the eight domains with only Connection to Nature below average and
that score did not differ significantly across groups. A broader analysis of value to human beneficiaries
like HWB tells a more complex story than a simple economic examination of community resources and
may indicate a more complex relationship between resource dependence and well-being. In point of fact,
a change in the weighting scheme (RIVs) (Smith et al. 2013a) used to calculate overall well-being could
alter the relationship altogether. An examination of these 'relative importance values' by community
type is ongoing and may bear important fruit for an understanding of human well-being in smaller rural
counties.
The way individual communities ranked the eight domains of the HWBI gives us useful information on
measuring success using local priorities. Education and Safety were consistently important as domains
of HWB. Other domains that were more variable in importance across communities, such as Social
Cohesion and Health, may be more dependent on current local conditions and therefore be more
susceptible to change through time. This does not make them less informative regarding HWB but
perhaps indicates they are less important for measuring long-term sustainability and more useful for
measuring short-term improvements. The importance of Education and Safety were split somewhat
among community classification categories and this is informative in that community classification has
proven informative regarding community values and dependencies (Chapter 2). The unweighted HWBI
also differs between classification groups with large differences in local resource dependence and these
differences are compounded by apparent differences in HWB priorities. All classification categories
were not fully evaluated with regards to RIVs and this should be a priority moving forward as present
data strongly suggest the CCS delineation may be useful for localizing the HWBI in a repeatable way
nationwide.
Shifts in the HWBI score as a function of domain weightings is an indication of where community
priorities do not match up with community performance. The domain rankings were based on
stakeholder input and represent a key element of 'localizing' the HWBI in a repeatable way. In most
cases the community priorities lined up well with objective indicators of HWB across the domains,
however in cases where this was not evident, such as St. Landry Parish, LA, it must be considered a
prime target for sustainability planning at the community level. In the case of St. Landry Parish,
workshop outcomes indicated Education is a high priority to the community, while indicators suggest
that it is an area of relatively weak performance in comparison to other domains of HWB. A focus on
education planning combined with a HWBI indicator RIV score emphasizing Education would allow for
both strong improvement in local HWB and an effective measure of when this had been achieved.
The HWBI also has the potential to serve as a measure of sustainable human well-being when tracked
through time and linked to alternative decisions that change the ecological, economic, and social states
of defined populations (Summers et al. 2014). However, the trajectories may change as a function of
community characteristics. For instance, the well-being of coastal communities is vulnerable to a variety
of economic, environmental and social factors. These vulnerabilities may be highlighted based on
community classification, particularly following episodic events such as hurricanes. Where affluence is
a strong characteristic of a community, the ability of that community to rebound economically following
such events may be much stronger than the abilities of less affluent areas, especially if local resource
dependence is low. However, the lack of Social Cohesion in combination with higher population
densities may result in an inequitable distribution of the restoration of services provisioning in those
areas. Therefore, if only viewed from an economic perspective, community well-being in affluent
communities may superficially seem to recover faster than in less affluent areas. On the other hand,
those rural communities affected by disasters, where Social Cohesion is strong, may rely on the
availability of human capital and require fewer external resources in order to rebound. These
27
-------
communities may reflect the same level of well-being as the affluent communities, but well-being is
derived from different levels of input and potentially more equitable (Cutter et al. 2008, Khazai et al.
2013).
3.4 Conclusions
Community decision makers can use the CCS to help identify baseline HWBI values from which to
assess the impact of decisions as shifts in community-specific HWB. Measures of community-specific
HWB also allow communities to restore, achieve, and sustain what matters most to them in terms of
human well-being. Connecting HWB measures to specific service flows, particularly environmental
service flows, is challenging, but the application of the CCS developed here can inform such a
connection by tying service flows directly to the social and economic characteristics of the community.
The environmental components (Ecoregion) of the CCS were less informative about community type
than the economic and social components (i.e., Lifemode and Location Quotient), yet the differences in
community type were strongly driven by economic and social dependence on local environmental
resource either through employment or through land use. This finding points to a clear link between
environmental service flows and HWB.
The approach of setting local HWB reference points based on community classification assumes that
common ground is important for describing community priorities. The limitations of this approach are
that specific factors important to individual communities are not considered and are likely to change in
importance across communities and at different spatial scales than considered here. Decision makers
wishing to set reference points for HWB will need to consider the consistency of the group assignments
to their situation, but in cases where this is an effective approach, much will be gained by allowing
similar communities to compare their HWBI values.
The community classification system developed during this study was also intended to inform decision
makers about a community's priorities. The association of these priorities with human well-being is a
tool for informing decision makers about sustainable decision outcomes in a community-specific
context. Overall, the CCS is useful for delineating coastal counties with respect to the three tiers of
sustainability. Local decision makers can identify a community's type based on CCS and then will have
a reference HWBI value from which to measure change in the HWBI, as well as information on what
dependencies are most important for their community. The limitations of working at the county level
need to be addressed as community characteristics are frequently very different at the county and local
scale. That said, the county scale is a natural starting place with clear linkages to local decision making.
The level of information about human well-being contained in the CCS appears promising for future
work and the CCS can be a useful tool for the purpose of setting needed HWBI benchmarks, as well as
gauging the transferability of sustainable decisions among coastal communities of similar type.
28
-------
3.5 Literature cited
Abunge, C., S. Coulthard, and T.M. Daw. 2013. Connecting marine ecosystem services to human well-
being: Insights from participatory well-being assessment in Kenya. Ambio 42:1010-1021.
Adeel, Z. and U. Safriel. 2008. Achieving sustainability by introducing alternative livelihoods.
Sustainability Science 3:125-133.
Butler, J.R.A., G.Y. Wong, D.J. Metcalfe, M. Honzak, P.L. Pert, N. Rao, M.E. van Grieken, T. Lawson,
C. Bruce, F.J. Kroon, and J.E. Brodie. 2013. An analysis of trade-offs between multiple
ecosystem services and stakeholders linked to land use and water quality management in the
Great Barrier Reef, Australia. Agriculture Ecosystems &Environment 180:176-191.
Canadian Index of Wellbeing. 2012. How are Canadians Really Doing? Canadian Index of Wellbeing
and University of Waterloo, Waterloo, ON.
Carriger, J.F., W.S. Fisher, T.B. Stockton Jr., and P.E. Sturm. 2013. Advancing the Guanica Bay (Puerto
Rico) watershed management plan. Coastal Management 41:19-38.
Condie, S.A., D. Hayes, E.A. Fulton, and M. Savina. 2012. Modelling ecological change over half a
century in a subtropical estuary: Impacts of climate change, land-use, urbanization and freshwater
extraction. Marine Ecology Progress Series 457:43-66.
Cutter, S.L., L. Barnes, M. Berry, C. Burton, E. Evans, E. Tate, and J. Webb. 2008. A place-based model
for understanding community resilience to natural disasters. Global Environmental Change-
Human and Policy Dimensions 18:598-606.
Garcia-Llorente, M., B. Martin-Lopez, S. Diaz, and C. Montes. 2011. Can ecosystem properties be fully
translated into service values? An economic valuation of aquatic plant services. Ecological
Applications 21:3083-3103.
Gilbert, A. 1999. Urbanization and Security, Project on Urbanization, Population, Environment, and
Security. Woodrow Wilson International Center for Scholars, Washington, DC.
Grabowski, J.H., R.D. Brumbaugh, R.F. Conrad, A.G. Keeler, J.J. Opaluch, C.H. Peterson, M.F. Piehler,
S.P. Powers, and A.R. Smyth. 2012. Economic valuation of ecosystem services provided by
oyster reefs. Bioscience 62:900-909.
Green, A.L., L. Fernandes, G. Almany, R. Abesamis, E. McLeod, P.M. Alino, A.T. White, R. Salm, J.
Tanzer, and R.L. Pressey. 2014. Designing marine reserves for fisheries management,
biodiversity conservation, and climate change adaptation. Coastal Management 42:143-159.
Gregory, R.S. and R.L. Keeney. 2002. Making smarter environmental management decisions. Journal of
the American Water Resources Association 38:1601-1612.
Israel, B.A., A.J. Schulz, E.A. Parker, and A.B. Becker. 1998. Review of community-based research:
Assessing partnership approaches to improve public health. Annual Review of Public Health
19:173-202.
Isserman, A.M. 1977. The location quotient approach for estimating regional economic impacts. Journal
of the American Institute of Planning 43:33-41.
29
-------
James, A. and D. Aadland. 2011. The curse of natural resources: An empirical investigation of US
counties. Resource and Energy Economics 33:440-453.
Jorgenson, A.K., A. Alekseyko, and V. Giedraitis. 2014. Energy consumption, human well-being and
economic development in central and eastern European nations: A cautionary tale of
sustainability. Energy Policy 66:419-427.
Khazai, B., M. Merz, C. Schulz, and D. Borst. 2013. An integrated indicator framework for spatial
assessment of industrial and social vulnerability to indirect disaster losses. Natural Hazards
67:145-167.
Krotscheck, C., F. Konig, and L. Obernberger. 2000. Ecological assessment of integrated bioenergy
systems using the Sustainable Process Index. Biomass & Bioenergy 18:341-368.
Likens, G.E., C.T. Driscoll, and D.C. Buso. 1996. Long-term effects of acid rain: Response and recovery
of a forest ecosystem. Science 272:244-246.
Nelson, E.J., P. Kareiva, M. Ruckelshaus, K. Arkema, G. Geller, E. Girvetz, D. Goodrich, V. Matzek, M.
Pinsky, W. Reid, M. Saunders, D. Semmens, and H. Tallis. 2013. Climate change's impact on key
ecosystem services and the human well-being they support in the US. Frontiers in Ecology and
the Environment 11:483-493.
Neset, T.S.S. and D. Cordell. 2012. Global phosphorus scarcity: Identifying synergies for a sustainable
future. Journal of the Science of Food and Agriculture 92:2-6.
NRC. 2011. Sustainability and the U.S. EPA. The National Academies Press, National Research Council,
Washington, DC.
O'Higgins, T.G., S.P. Ferraro, D.D. Dantin, S.J. Jordan, and M.M. Chintala. 2010. Habitat scale mapping
of fisheries ecosystem service values in estuaries. Ecology and Society 15 p.
Pascual, M., A. Boija, J. Franco, D. Burdon, J.P. Atkins, and M. Elliott. 2012. What are the costs and
benefits of biodiversity recovery in a highly polluted estuary? Water Research 46:205-217.
Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F.T. Jr. 1998. Fishing down marine food webs.
Science 279:860-863.
Peterson, C.H., R.T. Kneib, and C.A. Manen. 2003. Scaling restoration actions in the marine environment
to meet quantitative targets of enhanced ecosystem services. Marine Ecology Progress Series
264:173-175.
Piazza, B.P., M.K. La Peyre, and B.D. Keim. 2010. Relating large-scale climate variability to local
species abundance: ENSO forcing and shrimp in Breton Sound, Louisiana, USA. Climate
Research 42:195-207.
Putzhuber, F. and H. Hasenauer. 2010. Deriving sustainability measures using statistical data: A case
study from the Eisenwurzen, Austria. Ecological Indicators 10:32-38.
Smith, L.M., H.M. Smith, J.L. Case, and L. Harwell. 2012. Indicators and methods for constructing a
U.S. human well-being index for ecosystem services research. U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-12/023.
30
-------
Smith, L.M., J.L. Case, L.C. Harwell, H.M. Smith, and J.K. Summers, 2013a. Development of relative
importance values as contribution weights for evaluating human wellbeing: An ecosystem
services example. Human Ecology 41 p.
Smith, L.M., J.L. Case, H.M. Smith, L.C. Harwell, and J.K. Summers. 2013b. Relating ecosystem
services to domains of human well-being: Foundation for a US index. Ecological Indicators
28:79-90.
Summers, J.K., L.M. Smith, J.L. Case, and R.A. Linthurst. 2012. A review of the elements of human
well-being with an emphasis on the contribution of ecosystem services. Ambio 41:327-340.
Summers, J.K., L.M. Smith, L.C. Harwell, J.L. Case, C.M. Wade, K.R. Straub, and H.M. Smith. 2014.
An index of human well-being for the U.S.: A TRIO approach. Sustainability 6:3915-3935.
Tallis, H., P. Kareiva, M. Marvier, and A. Chang. 2008. An ecosystem services framework to support
both practical conservation and economic development. Proceedings of the National Academy of
Sciences of the United States of America 105:9457-9464.
U.S. EPA. 2015. Community engagement for sustainability workshops report. RESES Project Report,
U.S. Environmental Protection Agency, Office of Research and Development, Washington, DC.
Vaissiere, A.C., H. Levrel, C. Hily, and D. Le Guyader. 2013. Selecting ecological indicators to compare
maintenance costs related to the compensation of damaged ecosystem services. Ecological
Indicators 29:255-269.
Velasquez, J.D., G. L'Huillier, L. Hernandez, V. Rebolledo, and M. Diaz. 2011. An information system
for monitoring the Chilean salmon industry sustainability. Ocean & Coastal Management 54:231-
240.
Yang, W., T. Dietz, D.B. Kramer, X. Chen, and J. Liu. 2013. Going beyond the millennium ecosystem
assessment: An index system of human well-being. Plos One 8 p.
Zar, J.H. 2010. Biostatistical Analysis, 3rd edition. Prentice Hall, New Jersey.
31
-------
4 Stakeholder-derived priorities
A key element of community decision making is to establish priorities and set measures of success. In
many cases this process is limited to a particular issue of current importance. For instance, a community
interested in increasing downtown walkability might choose the mean number of people observed
walking per hour as a measure of success. However, many communities have made significant
investments in strategic planning in an effort to achieve broader goals that extend beyond the current
issues of concern (e.g., Sustainable Communities; accessed 16 September 2016). These efforts present
an opportunity to compare goals across communities in an effort to find common ground. In this chapter,
two approaches are considered for identifying and classifying community priorities. The first approach
is direct engagement with a representative set of stakeholders in select communities to define their
fundamental objectives. Direct engagement with community stakeholders is more efficient for
identifying stakeholder objectives, but is vulnerable to criticism when the results are generalized either
to the entire community or to other communities. These vulnerabilities are addressed here and an
attempt is made to generalize the outcomes by using a structured approach for engagement. The second
approach is a more objective examination of strategic planning documents based on the identification
keywords associated with pre-defined priority categories. This approach requires fewer resources and
relies on the communities own efforts to set and define community priorities. However, the keyword
approach is also subject to criticisms that keywords have to be pre-selected and that the documents
analyzed do not follow a consistent format. An examination of community priorities will benefit from a
merging of the two approaches. The research goal here is to define community-specific priorities that
can be tied to environmental resources in later chapters to create useful measures of management
success. The goal within this chapter is to use stakeholder input to identify and rank community
priorities in a useful and consistent manner.
4.1 Direct stakeholder engagement
4.1.1 Introduction
From August 2014 through June 2015, the U.S. Environmental Protection Agency, Office of Research
and Development, Gulf Ecology Division (GED) conducted four Community Engagement for
Sustainability Workshops. The workshops focused on identifying stakeholder priorities and focused on a
'central issue' of concern identified by the communities themselves (e.g., downtown development).
Workshops were conducted in Pensacola, Florida; Thibodaux, Louisiana; Vero Beach, Florida; and
Opelousas, Louisiana. Table 4.1 describes the four communities, dates on which the workshops were
held, the number of community participants in each workshop, and the central issues addressed. Figure
4.1 identifies the four communities on a map. These workshops are a subset of those described in
Chapter 3.1.3. These communities were chosen based on the Community Classification System (Chapter
2) and targeted for more in depth analysis of results. Pensacola, FL, and Opelousas, LA, are both in CCS
Group 1 counties and Vero Beach, FL, and Thibodaux, LA, are both in Group 3 counties (Table 2.1)
providing a CCS comparison, as well as a comparison between two states for this analysis.
32
-------
Opelousas, LA
Pensacola, FL
Vero Beach, FL
Thibodaux, LA
Figure 4.1 Communities participating in community engagement for sustainability workshops.
Table 4.1 Workshop communities and central issues.
Community
Workshop Date
Number of
Participants
Central Issue
Pensacola, Florida
August 27, 2014
43
Strengthening existing neighborhoods for
a more vibrant Pensacola
Thibodaux, Louisiana
October 30, 2014
36
Sustaining Thibodaux's quality of life -
now and for the future
Vero Beach, Florida
February 24, 2015
32
Helping shape the growth and character of
Vero's downtown
Opelousas, Louisiana
June 11, 2015
30
Helping re-energize Opelousas'
recreational environment
Table 4.2 Key design elements and relationships to workshop objectives.
Design Element and Function
Objectives*
Pre-Workshop Planning and Organization
Pre-workshop demographic analysis: review of social, economic, housing and
other demographics to inform participant list
~
~
Pre-workshop conference calls: engage community leaders as partners to
support workshop; discuss expectations, agenda, and participation
~
~
Work with community to develop participant list: provide advice and assist
community in conducting broad, representative outreach
~
~
Development of central issue: identify issue that is salient, will compel broad
participation and will work with structured deliberative approach
~
~
~
~
33
-------
Design Element and Function
Objectives*
Workshop Design
Structured deliberative process: proceed through structured approach from
open-ended discussions through prioritization and ranking exercises; allow
adequate time for participants to understand goals; discuss basis for priorities
including trade-offs, short-term influences and goals "taken for granted"
~
~
~
Workshop flow and design: conclude values discussion before discussing more
narrow central issue, and tie-in to goals hierarchy at end
~
~
~
Central issue: include time for discussion of central issue to ensure usefulness of
workshop to community; use discussion to validate goals hierarchy
~
~
Workshop Materials
Community-facing agenda: create accessible agenda to help communicate
purpose, encourage participation and organize workshop flow
~
~
~
~
Goals starting-point document: translate HWBI indicators into structured goals
(direction-value-context) using simple language and examples; ensure that
participants understand goals and process accurately captures values
~
~
Domain ranking worksheet: reproduce questionnaire used to develop HWBI RIVs
to provide direct input for HWBI research
~
~
~
Workshop presentation: use presentation to facilitate workshop discussion and
flow and support participant understanding of instructions and concepts
~
~
~
Workshop Facilitation
Introduction and tone-setting: create "safe" environment with welcome from
local leader, icebreakers and ground rules to encourage full participation
~
~
~
Instructions: clearly describe activities and ensure participant understanding;
ensure that exercises and data reflect intentions
~
~
~
~
Active facilitation of break-out groups: monitor discussions and ensure that all
participants have input; answer questions about exercises
~
~
~
~
Active facilitation of large group discussions: use techniques to encourage broad
input, avoid pitfalls, clarify input and capture "community" perspective
~
~
~
~
Record and collect information: use/collect flipcharts, collect worksheets, take
notes and produce an accurate and complete record of community input
~
~
~
~
Post-Workshop Analysis and Reporting
Analyze workshop data: document workshop data, quantify rankings, link
central issue to long-term goals and produce summaries
~
~
~
34
-------
Design Element and Function
Objectives*
Produce community report: document workshop data, summarize workshop
findings and provide recommendations to encourage use of findings
~
~
~
Identify sustainability indicators: identify indicators most relevant to short- and
long-term goals; address community expectations
~
* Key
Val:
Validity or Accuracy: extent to which the process measures and accurately conveys community values,
goals, and priorities
Rel:
Reliability or Comparability: extent to which the process will produce consistent results when applied in
different contexts
Rep
Representativeness: degree to which data represent characteristics of the entire community population
(necessary but not sufficient for validity)
Co
Completeness: extent to which enough information is gathered to achieve the research objective
4.1.2 Overall goals
The Community Engagement for Sustainability Workshops were designed to achieve the following:
• Collect information on community priorities to be used for the following:
o Understand relationships between community priorities and available ecosystem goods
and services.
o Understand relationships between community priorities and human well-being.
o Understand relationships between community priorities and sustainability indicators that
could be useful to inform community decisions and monitor progress toward
sustainability.
• Provide practical advice to communities on the development of sustainability indicators that are
tailored to their core community priorities and critical needs.
• Demonstrate a workshop approach for stakeholder engagement based on a structured deliberative
process that is applicable in multiple communities.
4.1.3 Methods
Workshop design, implementation, and participation
The workshops were designed to meet the following data quality objectives: validity, reliability,
representativeness, and completeness. Critical to all of these objectives were efforts to ensure that a
broad range of community voices were included in a meaningful way in the discussion of community
values. The following subsections summarize the workshop design and implementation, as well as
participation outcomes.
35
-------
Overview of workshop design
The workshops were designed to produce reliable information about community priorities. The
workshop design was developed based on the structured decision making approach (Gregory and
Keeney 2002, Carriger et al. 2013), modified to account for the limited time available for each
workshop. Each of the workshops was split into two distinct parts. The first part focused on exploring
core community values using a series of structured discussions and exercises. A list of goals reflecting
the components of human well-being (Smith et al. 2012) was used to frame these activities and enable
comparison across workshops. The second part of each workshop focused on the community-defined
central issue and was an opportunity to apply community priorities to a practical question (e.g.,
downtown development).
This structured approach was intentionally designed to introduce the participants to categories of
community priorities in a stepwise fashion using non-technical language applied in a familiar context.
Part 1 started with open ended discussions of important community qualities. Participants were then
asked to "map" these qualities to the list of goals, which allowed the project team to capture the
discussions in a structured way while providing participants with a solid working knowledge of the well-
being categories. Subsequent exercises and discussions built on this foundation and asked participants to
prioritize goals and apply the resulting list of priorities to a central issue. The order of the day was
designed to encourage a broad discussion of core community values before narrowing the focus to the
central issue. Appendix A presents the standard workshop agenda. Appendix B presents the list of the
categories of human well-being derived that was used to structure workshop discussions and exercises.
Workshop implementation
The following subsection describes the different phases of the workshop. Each workshop followed the
same design and the workshops were facilitated by the same team from SRA International, Inc. to ensure
consistency and comparability of results. All workshops were conducted in a single day.
Pre-workshop planning and site visit
In the period leading up to each workshop, the facilitation team collaborated with local leaders to
identify the central issue, create a representative list of workshop participants, develop the workshop
invitation, and select an appropriate venue for the workshop. EPA encouraged the local leaders to
identify and define a central issue that would encourage broad participation and would benefit from a
structured discussion of community priorities. Local leaders were also encouraged to reach out and
invite community leaders and members who, as a group, would provide a representative perspective on
local community values.
The facilitation team met with local leaders multiple times prior to the workshop to assure workshop
goals of representativeness were achieved.
Preliminary workshop activities
As participants arrived to the workshops, they were asked to sign in and identify their affiliation with a
community organization or their self-described role as a member of the community. Each workshop
started with welcoming remarks from a local leader followed by introductions, a review of the agenda,
and a discussion of ground rules. Following this, each workshop followed the two-part agenda already
described.
36
-------
Workshop part 1: Building the foundation
As an "ice-breaker," the facilitators asked participants to describe their community as if they were
meeting someone new to the community, thinking about what makes the community unique. After this
open-ended discussion, participants worked in small groups of five to seven to identify the qualities of
the community that they cared about most.
Participants were then provided with the list of goals developed based on the domains and indicators
used in the HWBI (Smith et al. 2013) and were given time to review and ask clarifying questions about
the goals. In small groups, participants were asked to map the qualities identified in the previous
discussion to specific categories of goals (corresponding to HWBI categories).
Facilitators explained that goals associated with a community quality could be identified by asking
"why" the quality was important to them. Facilitators explained that in some cases an iterative series of
"why" questions might be needed to identify a goal and that more than one goal may apply. As an
example, facilitators explained that walkable streets could be identified as important because they
support health (physical activity), social cohesion (interactions with neighbors), safety and security
(feeling safe), or a combination of these goals. Following the mapping exercise, each group reported out
to the larger group.
Using a multi-voting (dot voting) process, each participant was given the opportunity to identify the
goals most important to him or her. Facilitators placed flip charts with the lists of goals on a wall in the
room and participants were each given seven dots. Participants were asked to place their dots next to the
goals that they felt were most important to the community. Participants were allowed to distribute the
seven dots next to one or more goals, placing as many dots next to a goal as desired.
The dot voting exercise was followed by a discussion with the large group about whether there were any
surprises and other observations about the results. Participants were also asked whether during the
mapping and voting exercises they identified any important goals that they felt were missing from the
list.
As a final exercise for the first part of the workshop, participants were asked to rank goal categories
(e.g., Health, Education) based on their individual views of how important each category is to the well-
being of members of the community. Participants were asked to use a scale from 1 (most important) to 8
(least important) with no ties.
During large group discussions, facilitators used active techniques to encourage broad input, avoid
pitfalls, clarify input and capture as broad a perspective as possible. Facilitators also monitored small
group discussions and, when necessary, engaged the group to ensure that the discussions were focused
and that all participants had the opportunity to provide input.
Workshop part 2: Central issue
During the second part of the workshop, participants discussed the community-specific central issue.
These discussions started with an introduction by a community leader to frame the issue. Subsequent
discussions were tailored to the specific needs of each community. The discussions generally involved
brainstorming ideas for addressing the issue; identifying and prioritizing short- and longer-term actions
that the community could take; and providing participants with an opportunity to identify commitments
regarding how they would contribute to a solution.
37
-------
At the outset of these discussions, the facilitators explained how the earlier discussions provided insights
into the community's core values and suggested that participants draw on those insights when discussing
the central issue. Participants subsequently noted how the earlier discussions about important
community goals and values had helped them frame their discussions about the central issue.
Follow-up activities
Following each workshop, EPA collected the flip charts and ranking worksheets, calculated summary
results for the different exercises, and developed and delivered to each community a workshop report.
Each report documented group discussions about community values, breakout exercises, dot voting and
ranking results, and discussions exploring the central issue. The workshop reports provided information
to help each community interpret workshop results in terms of core community values. Each report also
provided recommendations for using the results to help guide community decisions and actions to
address the central issue and other issues facing the community.
Workshop participation
A critical element of the design of the workshop was the selection of participants to best represent the
community. This affects the extent to which the information gathered in the workshops accurately
represents the values, goals and priorities of the community as a whole. The project team sought to
maximize representativeness of workshop participants in the planning and implementation of the
workshops.
During the planning phase for each workshop, the project team collaborated with local leaders to
identify, reach out, and secure the commitment of as diverse and representative a group of participants
as possible. The project team recognized that some members of a community would not feel comfortable
participating in this type of public forum and others would be precluded participating due to workshop
timing (i.e., during a normal workday). To help address this, community leaders were encouraged to
recruit participants to represent the voices of groups of community members who were less likely to
participate or unable to attend. The team emphasized that in addition to public officials and others who
normally participate in these type of forums, "community leaders" include others that are well-
connected and respected within different parts of the community, such as church leaders and
neighborhood association officers.
The facilitation team also asked participants to "wear different hats" during the workshop and try to
represent not only their own viewpoints but also those of community members who could not attend.
Participants were asked to report their affiliations when signing in at the start of the workshops. Table
4.3 summarizes the self-reported affiliations of workshop participants.
38
-------
Table 4.3 Workshop participation for major interest groups by community. Number of persons attending each
workshop given in parentheses.
Affiliation
Opelousas
(30)
Pensacola
(44)
Thibodaux
(36)
Vero Beach
(33)
City leadership
~
~
~
~
Local business owners
s
s
s
s
Civic organizations
s
s
s
s
Neighborhood associations
s
s
Other social organizations
s
s
Church leadership
s
s
Regional planning organizations
s
s
s
City agencies
s
s
s
s
Educators (K-12 and college)
s
s
s
s
Young professionals
s
s
s
s
Students (secondary, college)
s
Retirees
s
s
s
s
Table 4.4 Characteristics of four study communities. Data are given separately for change overtime (2000-
2013) in census data and the results of the American community survey (ACS) for the five year period ending in
2013.
Characteristic
Opelousas
Pensacola
Thibodaux
Vero Beach
Change
2013
Change
2013
Change
2013
Change
2013
2000 to
5-yr
2000 to
5-yr
2000 to
5-yr
2000 to
5-yr ACS
2013
ACS
2013
ACS
2013
ACS
2013
People and Households
Population demographics
Total population
-27%
16,679
-7%
52,268
1%
14,576
-13%
15,475
Total households
-32%
5,927
-10%
22,150
-2%
5,400
-14%
7,312
Share of households that are
2%
67%
-7%
56%
-5%
58%
-12%
49%
family households
Share of households with
1%
39%
-19%
23%
-13%
29%
-4%
19%
children and youth
Share of population aged 20
-3%
18%
10%
20%
8%
26%
-13%
14%
to 34 years
Share of population 65 years
-11%
14%
1%
17%
2%
14%
-8%
27%
and older
Generational mixing
18%
2.25
-15%
1.13
-21%
1.37
12%
0.61
Age diversity
0%
0.91
0%
0.91
0%
0.91
0%
0.91
Ethnic and racial diversity
-16%
0.37
3%
0.52
6%
0.51
27%
0.27
Factors affecting socioeconomic status
Share of population that are
27%
73%
7%
91%
11%
77%
2%
86%
high school graduates or
higher
Share of population with
6%
12%
4%
34%
6%
23%
0%
32%
bachelor's degree or higher
Share of younger population
-50%
5%
-16%
30%
-2%
26%
-56%
12%
that are college graduates
Share of population with
NA
81%
NA
81%
NA
83%
NA
79%
health insurance coverage
39
-------
Characteristic
Opelousas
Pensacola
Thibodaux
Vero Beach
Change
2013
Change
2013
Change
2013
Change
2013
2000 to
5-yr
2000 to
5-yr
2000 to
5-yr
2000 to
5-yr ACS
2013
ACS
2013
ACS
2013
ACS
2013
Share of population with a
NA
14%
NA
15%
NA
16%
NA
17%
disability
Population stability
Share of population in the
NA
86%
NA
82%
NA
81%
NA
77%
same house as 1 year ago
Housing unit vacancy rate
42%
16%
54%
14%
38%
12%
68%
29%
Share of housing units that
-9%
49%
-6%
59%
8%
56%
-3%
63%
are owner-occupied
Share of occupied housing
1%
47%
5%
38%
-14%
41%
2%
35%
that are rental units
Population Density and Housing
Population density (people
-27%
2,110
-7%
2,319
1%
2,419
-13%
1,353
per square mile)
Number of housing units
-28%
7,034
-4%
25,797
2%
6,108
0%
10,286
Share of housing built 1990
NA
16%
NA
15%
NA
26%
NA
20%
or later
Share of housing built before
NA
7%
NA
11%
NA
10%
NA
5%
1940
Median value of owner-
39%
$75.IK
62%
$151.3
106%
$148.6
32%
$191.8K
occupied housing units
K
K
Owner-occupied housing
NA
0.75
NA
0.83
NA
0.83
NA
0.85
value diversity
Economy
Employment and income
Labor force participation rate
7%
48%
-2%
64%
-2%
58%
-1%
52%
Unemployment rate
-22%
13%
67%
10%
-39%
5%
317%
14%
Median household income
37%
$20,16
27%
$44,14
61%
$43,05
-4%
$37,051
5
4
8
Average household income
NA
$34,52
NA
$62,68
NA
$62,63
NA
$67,715
9
0
8
Household income diversity
1%
0.82
0%
0.82
-1%
0.82
1%
0.82
Gini index of income
NA
0.52
NA
0.48
NA
0.50
NA
0.57
inequality
Share of families in poverty
91%
37%
94%
12%
-23%
9%
91%
12%
Share of individuals in
57%
42%
91%
17%
-1%
16%
124%
20%
poverty
Affordability
Share of owner-resident
171%
29%
27%
27%
61%
11%
29%
35%
households with housing
costs > 35% of income
Share of renting households
34%
53%
22%
43%
9%
37%
91%
59%
with gross rent > 35% of
income
Local economy
Share of workforce in
-16%
22%
-30%
37%
-8%
30%
-4%
32%
management, professional,
and related occupations
Share of workforce in service
25%
31%
41%
20%
8%
20%
15%
23%
40
-------
Characteristic
Opelousas
Change 2013
2000 to 5-yr
2013 ACS
Pensacola
Change 2013
2000 to 5-yr
2013 ACS
Thibodaux
Change 2013
2000 to 5-yr
2013 ACS
Vero Beach
Change 2013
2000 to 5-yr ACS
2013
occupations
Share of workforce in sales
and office occupations
12%
24%
36%
28%
-9%
25%
12%
31%
Share of workforce in natural
resources, construction, and
maintenance occupations
-13%
9%
31%
9%
15%
10%
-8%
11%
Share of workforce in
production, transportation,
and material moving
occupations
-17%
14%
-21%
7%
19%
14%
-50%
4%
Share of commuting
workforce with travel time >
45 minutes
9%
11%
-4%
7%
6%
12%
7%
7%
Share of workforce who
worked in county of
residence
-13%
77%
2%
90%
-27%
65%
-5%
89%
During the workshops, the facilitation team asked participants to comment on groups that were not
adequately represented in the workshop. In all of the workshops, participants felt that lower income
residents were underrepresented. Specific discussions regarding these and other underrepresented
community members are summarized below:
• Pensacola, FL: Participants noted that two key challenges are the high school drop-out rate and
the poverty level within the community. Participants felt that those who could represent these
challenges first-hand (e.g., church leaders) were not well represented in the workshop.
Additionally, participants noted that the ability of younger people to see opportunities in
Pensacola was critical to the community's future. Participants felt that this segment of the
community had some, but insufficient representation.
• Thibodaux, LA: Participants noted that the relatively significant low income population of the
community was not well represented in the workshop.
• Vero Beach, FL: Participants noted that to sustain its vitality, the community will need to
provide greater opportunities and amenities for young adults and young families. Participants
discussed the belief that this group does not feel that they have an adequate voice in community
decisions. Participants agreed that this group was not well-represented in the workshop.
• Opelousas, LA: Participants noted that a key to their success was to engage trusted leaders that
represent the diverse age and socio-economic groups within the community. They believed that
the workshop would have benefited from more church leaders and leaders who could represent
different neighborhoods.
41
-------
4.1.4 Results
Summary data on index communities
This subsection summarizes background information for the four index communities that participated in
the workshops. It also provides a summary and comparison of the central issues addressed during the
workshops. Combined, this information can be used to:
• Identify key stressors being faced by these communities, where the background data on social
and economic details provide empirical evidence of social and economic stressors and the central
issues provide an indication of the community's perceptions, drawing on local knowledge and
experience.
• Conduct preliminary analyses of associations between socio-economic variables and workshop
outcomes to inform future research on relationships between measurable community
characteristics and the nature and structure of core community values.
Comparison of communities by socio-economic characteristics
To provide context for interpreting workshop results, socio-economic data were collected for the four
participating communities. Socio-economic variables were selected based on potential relevance to
community sustainability, including:
• Population demographics and trends, including the share of the population that is young
professionals and families, age distribution, and ethnic, racial, and socio-economic diversity.
• Indicators of community resilience, including population stability, disability, health insurance
coverage, economic diversity, and indicators of socio-economic status (e.g., educational
attainment, income, and employment status).
• Economic setting and trends, including trends in income, poverty, and affordability and the
nature of local economies and their dependence on natural resources.
Some of the more significant similarities and differences among the communities include:
• Total population — The population of Pensacola, FL, is three times larger than each of the other
communities, which have similar total populations. The communities have similar population
densities, except for Vero Beach, FL, which is less dense. Opelousas, LA, and Vero Beach, FL,
underwent significant population declines between 2000 and the five-year period ending in 2013.
• Population age — In the 2013 American Community Survey (ACS; American Community
Survey; accessed 16 September 2016), Opelousas, LA, had the highest concentration of
households with children and youth and highest degree of generational mixing (i.e., the ratio of
population under 17 and under to population 65 and over). Vero Beach, FL, has the lowest
concentration of households with children and youth and the lowest degree of generational
mixing. Pensacola, FL, and Thibodaux, LA, have most significant shares of population in 20 to
34 range.
• Ethnic and racial diversity — Opelousas, LA, and Vero Beach, FL, had lower levels of ethnic
and racial diversity than the other two communities. The majority of residents in Opelousas
identify themselves as Black or African American alone (not Hispanic or Latino) and the
majority of residents in Vero Beach, FL, identify themselves as White alone (not Hispanic or
Latino).
42
-------
• Housing occupancy and ownership — Housing vacancy rates increased in all four communities
from 2000 to the five-year period ending in 2013. Three of the four communities saw a decrease
in the share of owner-occupied housing and an increase in the share of renter-occupied housing
over this period. Thibodaux, LA, was the exception and experienced the opposite trend.
• Income and wealth — In the 2013 ACS, Opelousas, LA, had the lowest median home values,
lowest median and mean incomes, and highest shares of families and individuals in poverty.
Vero Beach, FL, had the third highest median income, highest mean income, highest GINI index
(income distribution equity; The World Bank GINI index; accessed 16 September 2016) and
highest housing value diversity indicating potential income and wealth disparities among
households in the community. All of the communities except Thibodaux, LA, experienced
increases in families and people in poverty from 2000 to the five-year period ending in 2013.
• Affordability — All of the communities experienced increases from 2000 to the five-year period
ending in 2013 in the share of homeowners and renters who spend more than 35% of income on
housing. In the 2013 ACS, Vero Beach, FL, had the highest share of owners and renters who
paid 35% or more of their income on housing costs.
• Occupations — The 2013 ACS data indicate that Opelousas, LA, had the lowest share of the
workforce in management, professional, and related occupations and highest share in service
occupations. Of the four communities, the two Louisiana communities had the highest
proportions of their workforce in production, transportation, and material moving occupations.
Workers in the Louisiana communities also had the longest commutes, suggesting employment
outside of their resident communities.
Presentation and comparison of central issues
The central issue was selected by each community based on its prominence as an issue being faced by
the community and likelihood to resonate with and attract a representative group of participants. The
choice of a central issue could tend to bias participation in a workshop by attracting people with a deeper
interest in the issue. Therefore, it is important account for the choice of the central issue when
interpreting workshop outcomes. The central issue could also be an indicator of the stressors and
opportunities being faced by a community, with relevance for identifying meaningful community-
specific sustainability indicators. The following subsections identify similarities and differences in the
central issues addressed during the workshops.
Similar themes
While each of the central issues addressed distinct and different aspects of the community, there were
some similar themes that emerged across the workshops:
• All of the communities believed that by addressing the central issue, a key outcome that will be
achieved is improved Social Cohesion; building connections across the community that
ultimately will support longer term resilience.
• Preserving or strengthening ties to the history and culture of the community was another key
outcome related to each of the central issues.
• While each community's actions varied, improving safety was another common theme.
• Pensacola, FL; Thibodaux, LA; and Vero Beach, FL, stressed the importance of the central issue
to the longer term economy of the community.
43
-------
Significant differences:
Key differences in focus among the central issues addressed in the workshops include:
• Pensacola, FL — In this workshop, there was strong emphasis on Living Standards and the
importance of vibrant neighborhoods in supporting youth and on improving the conditions for
the least well off members of the community and instilling hope.
• Thibodaux, LA — While the central issue addressed many key goals, there was very little tie to
Education, despite the fact that Education was viewed as one of the most important values.
• Vero Beach, FL — A unique concept that emerged from this workshop was that Vero Beach, FL,
is a community without a clear identity. While participants were clear on who we "don't want to
be" there was not a clear vision of "who we want to be" that can guide the development of the
downtown.
• Opelousas, LA — The community is in the midst of a "rebuilding" era, still recovering from a
significant hit by the economy and with new energy that the new mayor is bringing. The central
issue can help the community overcome negative perceptions.
Comparison of index communities
Comparison of values mapping exercise
As discussed in Chapter 4.1.3, participants worked in small groups to identify the qualities of the
community that they cared about most. Participants were then asked to map the qualities identified in the
previous discussion to specific categories of goals corresponding to the HWBI domains. Table 4.5 and
Figure 4.2 summarize the frequency with which a goal category was identified by the small groups
during this exercise. Frequencies were calculated as the total number of times that a goal category was
associated with a community quality by any small group divided by the total instances that any goal
category was identified by any small group. For example, if the workshop included four small groups,
each group listed five important community qualities during the open-ended exercise, and each group
associated two goal categories with each quality, the total instances that goal categories were identified
would be 40 (4*5*2). If one group identified the "health" category with two community qualities,
another identified "health" with one quality, and none of the other groups identified health, the
workshop-level frequency with which the "health" goal category was identified would be calculated as
7.5% (3/40).
44
-------
30%
25%
I 20%
>•
U
| 15% -hi
o-
ai
o 10%
>
-------
Participants in the Opelousas, LA, and Vero Beach, FL, workshops identified work life balance more
frequently and Social Cohesion less frequently than participants in the Pensacola, FL, and Thibodaux,
LA, workshops. Participants in the Opelousas, LA, and Thibodaux, LA, workshops similarly identified
Cultural Fulfilment less frequently that participants in the workshops held in the two Florida
communities.
Comparison of dot voting exercise
Following the mapping exercise, participants voted for the goals that they felt were most important to
the community by placing dots next to the goals, as described in Chapter 4.1.3. For the purpose of this
comparative analysis, the project team summed the goals to develop a total for each goal category. Table
4.6 and Figure 4.3 present the percentage of dot votes received by goals within each goal category
during each workshop.
Table 4.6 Relative frequency of goal categories receiving dot votes during workshop group voting exercise.
Each attendee was given eight dots. Goal categories are identical to the domains of the human well-being index
(HWBI).
Community
Education
Health
Work Life
Balance
Living
Standards
Safety and
Security
Connection
to Nature
Cultural
Fulfillment
Social
Cohesion
Opelousas
17%
18%
14%
15%
12%
5%
4%
14%
Pensacola
24%
14%
4%
21%
13%
2%
4%
19%
Thibodaux
22%
12%
13%
14%
14%
2%
3%
20%
Vero Beach
16%
21%
15%
13%
8%
7%
7%
14%
25%
S S. 20%
I a
H- • u 15%
c re
.
£
5 £
£ ~ 5%
0%
t
u
IJI
Education Health Work Life Living Safety and Connection Cultural Social
Balance Standards Security to Nature fulfillment Cohesion
Goal Category
¦ Pensacola
¦ Thibodaux
Vero Beach
¦ Opelousas
Figure 4.3 Relative frequency of goal categories receiving dot votes during workshop group voting exercise.
Goal categories are identical to the domains of the human well-being index (HWBI).
Three of the four workshops identified Education as the most important goal category to the community
when goal-specific dot votes were summed to the category level. Education was identified as the second
most important goal category in the fourth workshop. Social Cohesion received the second or third
highest number of summed votes in all four workshops, and Health received the first, second or third
46
-------
highest number of summed votes in three of the workshops. Living Standards, Connection to Nature,
and Cultural Fulfillment ranked sixth, seventh, or eighth compared to other goal categories based on
summed dot votes in all four workshops. Significant differences among workshops were identified using
an outlier analysis. The summary describes differences where the outcomes of the workshop were at
least 1.4 standard deviations from the mean. Significant differences among workshops include:
• The Pensacola, FL, workshop stood out based on the relatively low number of votes that goals
associated with Work Life Balance received.
• The Thibodaux, LA, workshop stood out based on the relatively low number of votes that goals
associated with Living Standards received.
• The Vero Beach, FL, workshop stood out based on the relatively high number of votes for goals
associated with the Health, Connection to Nature, and Cultural Fulfilment and relatively low
number of votes for goals associated with Safety and Security.
Participants in the Pensacola, FL, and Thibodaux, LA, workshops cast a greater percentage of votes for
goals associated with Education and goals associated with Social Cohesion than participants in the
Opelousas, LA, and Vero Beach, FL, workshops. Community similarities in these rankings were
analyzed using a simple univariate cluster analysis. Workshops were ranked based on the relative
frequency that a goal category received dot votes. If the differences in frequency between the first and
second ranked workshops and between the third and fourth ranked workshops exceeded the difference
between the second and third ranked workshops by 10%, the workshops were considered clustered in
pairs.
Comparison of domain ranking exercise
As a final exercise before discussing the central issue, participants were asked to rank goal categories
(e.g., Health, Education) based on their individual views of how important each category is to the well-
being of members of the community. Participants used a scale from 1 (most important) to 8 (least
important) with no ties. Table 4.7 and Figure 4.4 summarize the likelihood that a workshop participant
ranked a goal category as first, second or third most important. The likelihood is calculated as the total
number of participants who ranked a goal category as first, second or third most important divided by
the total number of first, second or third ranking instances, where the total instances equals the number
of participants in the exercise times three.
Table 4.7 Likelihood a goal category was identified in top three during a workshop in an individual ranking
exercise. Goal categories are identical to the domains of the human well-being index (HWBI).
Community
Education
Health
Work Life
Balance
Living
Standards
Safety
and
Security
Connection
to Nature
Cultural
Fulfillment
Social
Cohesion
Opelousas
32%
21%
9%
11%
12%
2%
3%
11%
Pensacola
22%
18%
1%
13%
22%
0%
4%
21%
Thibodaux
26%
9%
5%
16%
22%
1%
1%
20%
Vero Beach
13%
17%
6%
16%
22%
6%
6%
14%
47
-------
35%
30%
25%
« 20%
15%
10%
5%
0%
lIlLiLifc
Education Health Work Life Living Safety and Connection Cultural Social
Balance Standards Security to Nature Fulfillment Cohesion
i Pensacola
¦ Thibodaux
Vero Beach
I Opelousas
Goal Category
Figure 4.4 Likelihood a goal category was identified as top three priority during a workshop in an
individual ranking exercise. Goal categories are identical to the domains of the human well-being index
(HWBI).
Education was the goal category most likely to be ranked as one of the top three most important
categories by individual participants in three of the four workshops. Safety and Security was the goal
category most likely to be ranked as one of the top three by participants in two of the four workshops.
Health was the second most likely category to be ranked in the top three in two workshops, and Social
Cohesion was the third most likely to be ranked in the top three in two workshops. Work Life Balance,
Connection to Nature, and Cultural Fulfilment were the among the three least likely goal categories to
be ranked in the top three by participants in all four workshops. Significant differences among
workshops were identified using an outlier analysis. The summary describes differences were the
outcomes of the workshop were at least 1.4 standard deviations from the mean.
Significant differences among workshops include:
• The Opelousas, LA, workshop stood out from the others based on the relatively low rankings
received for Safety and Security and Social Cohesion.
• The Pensacola, FL, workshop stood out based on the relatively low rankings received for Work
Life Balance.
• The Thibodaux, LA, workshop stood out based on the relatively low rankings received for
Health.
• The Vero Beach, FL, workshop stood out based on the relatively low rankings received for
Education and the relatively high rankings received for Connection to Nature.
Participants in the Thibodaux, LA, and Vero Beach, FL, workshops ranked Living Standards as
relatively more important than participants in the Opelousas, LA, and Pensacola, FL, workshops.
Workshop pairings were analyzed using a simple univariate cluster analysis. Workshops were ranked
based on the likelihood that a goal category was ranked as a first, second or third priority. If the
differences in likelihood between the first and second ranked workshops and between the third and
fourth ranked workshops exceeded the difference between the second and third ranked workshops by
10%, the workshops were considered clustered in pairs. Participants in the Pensacola, FL, and
48
-------
Thibodaux, LA, workshops ranked Social Cohesion as relatively more important than participants in the
Opelousas, LA, and Vero Beach, FL, workshops.
Comparison of mapping, dot voting, and ranking outcomes
Workshop data gathered using mapping, dot voting, and ranking exercises were compared to assess
consistency and complementarity — in terms of identifying core community values — across the
different types of exercises. When interpreting the comparative analysis, the project team considered the
following key differences among the exercises:
• The exercises involved different levels of structuring. The mapping exercise asked participants to
address an open-ended question (i.e., "what do you care about most in a community") and map
goals retrospectively. The other two exercises asked participants to prioritize a specific list of
goals and goal categories.
• The dot voting exercise asked participants to prioritize goals at the equivalent of the HWBI
"indicator" level of resolution. The other two exercises involved mapping and ranking of the
more aggregated goal categories, with a resolution equivalent to the HWBI "domain."
• The structured exercises (i.e., dot voting and ranking) differed in the anonymity afforded
participants. Dot voting was a community activity where participants could observe and react to
the choices of others and could expect others to observe their choices. The ranking activity was
completed by participants individually and anonymously.
• The workshop was designed to incrementally introduce participants to the goals framework and
concepts. When completing later exercises, participants benefited from a deeper understanding
of the material and improved capacity to express their intent.
For comparison, the results of the mapping, dot voting and ranking exercises were expressed in terms of
ranked goal categories. For example, the goal category that received the highest number of aggregated
votes in dot voting was ranked first out of the eight goal categories and that receiving the lowest number
was ranked eighth. A similar approach was used to rank goal categories based on frequency of
identification during the mapping exercise and likelihood of being ranked as one of the top three
categories in the ranking exercise. Figure 4.5 compares the ranking of the goal categories across the
different exercises by community.
49
-------
Education
Mapping Dot Voting Ranking
¦Opelousas Pensacola
Thibodaux Vero Beach
Work Life Balance
Mapping Dot Voting Ranking
¦Opelousas Pensacola
Thibodaux Vero Beach
Safety and Security
Mapping Dot Voting Ranking
¦Opelousas Pensacola
Thibodaux Vero Beach
Cultural Fulfillment
Mapping Dot Voting Ranking
¦Opelousas Pensacola
Thibodaux Vero Beach
Health
Mapping Dot Voting
Ranking
¦Opelousas Pensacola
Thibodaux Vero Beach
Living Standards
Mapping Dot Voting Ranking
¦Opelousas
Thibodaux
¦Pensacola
¦Vero Beach
Connection to Nature
Mapping Dot Voting Ranking
¦Opelousas Pensacola
Thibodaux Vero Beach
Social Cohesion
Mapping Dot Voting Ranking
¦Opelousas
Thibodaux
¦Pensacola
¦Vero Beach
Figure 4.5 Comparison of workshop outcomes across the three weighting exercises. Data are given separately
for workshops in each of the four communities.
50
-------
Key findings from comparing the workshop outcomes across the different exercises include:
• Education and Health goal categories were ranked consistently high across the three exercises. In
three communities, Education increased in ranking from mapping to dot voting and remained the
top ranked goal category based on the ranking exercise. In one workshop, Education was
identified as a relatively lower priority in the ranking exercise. From dot voting to ranking, the
Health category shifted by no more than one rank.
• Living Standards and Safety and Security goal categories followed a similar pattern. Their
relative rank based on the exercise-specific metrics decreased from mapping to dot voting and
increased from dot voting to ranking. In all four workshops, these two goal categories were
identified as relatively higher priorities during the ranking exercise than they were based on dot
voting.
• Work Life Balance and Social Cohesion were identified as relatively lower priorities during the
ranking exercise than they were based on dot voting.
• Connection to Nature and Cultural Fulfillment were associated relatively frequently with the
important community qualities identified at the outset of the mapping exercise. They were
identified as relatively lower priorities based on the dot voting and ranking exercises.
These preliminary findings and complementary observations made during the workshops suggest the
following:
• The structured deliberative process is a useful tool for revealing core community values. When
the goals framework was introduced, the focus of the workshops tended to shift from the natural
and cultural qualities that establish a community's identity to core concerns such as safety and
reasonable living standards.
• The ranking exercise may be a most accurate expression of core community values for the
following reasons:
o The exercise is less subject to the introduction of strategic considerations (e.g., trying to
spread votes evenly across goal categories).
o By the time participants complete the ranking exercise, they have a deeper understanding
of the goal category definitions and have had greater opportunity to form their thoughts
about priorities.
o The anonymity in the ranking exercise helps to control for response bias, as participants
are less likely to consider others' perceptions when expressing priorities.
• All three of the exercises provide useful and complementary information relative to a
community's core values and useful sustainability indicators. The mapping and dot voting
exercises complement the ranking exercise by revealing interrelationships between intermediate
and end goals, stressors driving community priorities and action, and unique qualities of a
community that can be leveraged to achieve end goals.
Analysis of socio-economic characteristics and workshop outcomes
Workshop and socio-economic data were analyzed to address questions about generalizing workshop
outcomes within and across communities. These included simple cluster and regression analyses tailored
to account for the limited number of workshops for which data are available.
51
-------
Simple univariate cluster analyses were conducted to identify community pairings based on similarities
in terms of socio-economic characteristics or workshop domain rankings. Bivariate cluster analyses
were also conducted to explore possible interactions between socio-economic factors and workshop
outcomes. Clustering based on each combination of one socio-economic factor and one workshop
outcomes was analyzed in bivariate space using similar decision criteria as the univariate analysis (i.e.,
ranked pairs with at least a 10% separation). The analysis concluded that clusters were determined by
either the socio-economic factor or the workshop outcome and that the bivariate analysis did not add to
the findings from the univariate analyses. For the socio-economic analyses, communities were ranked
based on each socio-economic factor. If the difference in the factor values between the first and second
ranked communities and between the third and fourth ranked communities exceeded the difference
between the second and third ranked communities by at least 10%, the communities were considered
clustered in pairs. The same approach was used to assess clustering of workshops based on workshop
outcomes. Table 4.8 summarizes the results of the univariate cluster analyses.
Table 4.8 Summary of univariate cluster analyses between communities.
Community Pairings
Observed Clustering: Clustered
on Socio-economic
Characteristics
Observed clustering: Clustered on
Goal Categories*
Opelousas - Pensacola
Thibodaux - Vero Beach
• Share of workforce in natural
resources, construction, and
maintenance occupations
• Living standards (R)
Opelousas - Thibodaux
Pensacola - Vero Beach
• Share of population that are
high school graduates or higher
• Share of workforce in
production, transportation, and
material moving occupations
• Share of commuting workforce
with travel time >= 45 minutes
• Cultural fulfillment (M)
Opelousas - Vero Beach
Pensacola - Thibodaux
• Ethnic and racial diversity
• Share of population that is
younger college graduates
• Share of housing built before
1940
• Share of renting households
with gross rent > 35% of income
• Education (V)
• Work life balance (M)
• Social cohesion (MVR)
* Labels in parentheses denote the source of workshop outcome data in terms of type of workshop exercise as follows: M =
mapping, V = dot voting, R = ranking.
Possible associations between socio-economic factors and workshop outcomes were explored using
bivariate linear regression analysis with strong decision criteria. Each workshop outcome was regressed
on each socio-economic factor using both the 2013 5-year ACS data and the change in value between
the 2000 Decennial Census (U.S. Census 2010; accessed 14 September 2016) and the 2013 five-year
ACS; and was tested for both consistency in rank and linearity (R2> 0.8). For example, the community
with the highest value for a workshop outcome (e.g., relatively frequency of dot votes for a goal
category) also had the highest value for a socio-economic factor, the community with the second highest
value for a workshop outcome also had the second highest value for a socio-economic factor, etc.
Inverse ranking relationships were also considered (i.e., highest-to-lowest corresponding to lowest-to-
52
-------
highest). This approach was selected based on the small number of communities and workshops
included in the analysis. It did not consider interactions among multiple socio-economic variables and
not intended to evaluate causality. Table 4.9 summarizes the results of the preliminary association
analysis.
Table 4.9 Summary of associations' analysis across communities.
Characteristic
Observed Associations between Socio-Economic Factors and
Workshop Outcomes*
Education
Health
Work Life
Balance
Living
Standards
Safety and
Security
Connection
to Nature
Cultural
Fulfillment
Social
Cohesion
People and Households
Population demographics
Share of households that are family households
P
(MR)
Share of households with children and youth
P
(MR)
Share of population aged 20 to 34 years
N
(MV)
P (V)
N (V)
Generational mixing
P
(MR)
Age diversity
P
(MR)
Ethnic and racial diversity
P (V)
Factors affecting socioeconomic status
Share of population that is younger college
graduates
P (R)
Population stability
Housing unit vacancy rate
P (V)
N (V)
P (V)
P (V)
Share of housing units that are owner-occupied
N
(M)
Share of occupied housing that are rental units
P
(MR)
Population Density and Housing
Population density (people per square mile)
N (V)
P (V)
N (V)
N (V)
Share of housing built 1990 or later
P (M)
Share of housing built before 1940
P (V)
P (M)
Median value or owner-occupied housing units
N
(M)
Economy
Employment and income
Labor force participation rate
P (R)
Unemployment rate
P
(MV)
Median household income
P (R)
Average household income
N
(M)
Household income diversity
P (R)
GINI index of income inequality
N (V)
N
(M)
P (Ft)
Affordability
Share of owner HH with housing costs > 35% of
income
P (M)
53
-------
Characteristic
Observed Associations between Socio-Economic Factors and
Workshop Outcomes*
Education
Health
Work Life
Balance
Living
Standards
Safety and
Security
Connection
to Nature
Cultural
Fulfillment
Social
Cohesion
Share of renting HH with gross rent > 35% of
income
P
(MV)
N (V)
P (V)
N (V)
Local economy
Share of workforce in sales and office occupations
N
(M)
Share of commuters with travel time > 45 minutes
N
(M)
Share of workforce who worked in county of
residence
P (M)
* Key: Direction of association: P = positive linear association, N = negative linear association. Source of workshop data: M =
mapping, V = dot voting, R = ranking
4.1.5 Discussion
Practical measures of sustainability
Key objectives of the Community Engagement for Sustainability Workshops were to provide practical
advice to communities and to inform EPA research on sustainability indicators. The following
subsections present the findings from the workshops relative to these objectives.
Description of spreadsheet tool for sustainability indicators
Measuring progress towards sustainability is a key challenge for communities. While a wealth of
"sustainability indicators" are available (Fang et al. 2014, Holden 2013, Vackar et al. 2012), the
challenge for many communities is deciding which sustainability indicators are most relevant to their
situation. Defining sustainability indicators in terms of the core community values identified through the
types of processes used in the workshops described herein offers a solution to this challenge. It allows
communities to select indicators focused on the values-oriented goals that the community wishes to
achieve or sustain and create a succinct and meaningful system of indicators tailored to these goals.
To demonstrate this approach, the project team identified community sustainability indicators that
resonated with the information received during the workshops, including information on core values,
associated near- and long-term goals, and strategies for achieving those goals. The team developed a
spreadsheet and navigation tool using the central issues addressed in each workshop to show how
communities could develop a set of interrelated indicators that is relevant to a specific issue and directly
linked to core values and associated goals.
The approach is based on four types of indicators (Figure 4.6), each of which has a different relationship
to factors affecting a community, community decisions and actions, and community well-being:
• Indicators of external factors affecting the community (Type 1) — Measures of forces affecting
community sustainability that are beyond a community's direct control (e.g., climate change,
54
-------
national economic trends). These indicators could be used to help communities understand the
root causes behind existing and emerging challenges, target strategies to adapt, and communicate
the need for action.
• Indicators of consequences of external factors (Type 2) — Measures of the effects of external
factors on the ability of a community to sustain or achieve its goals, including measures of
changes in conditions relative to intermediate goals and goals directly related to core values (e.g.,
highly ranked domains of the HWBI). These indicators vary by external factor and could be used
to provide the rationale for community action, set priorities, and guide community action.
• Indicators of possible community actions (Type 3) — Measures of the status and immediate
outcomes of community actions (e.g., acreage of greenspace added to a downtown). These
indicators could be used to establish milestones for action, report progress, and demonstrate
accountability to the public. These indicators recognize that immediate outcomes often depend
on responses of key stakeholders and provide the most immediate feedback to help communities
change course if an action is not working as intended.
• Indicators of the outcomes of community actions (Type 4) — These indicators align with
indicators of consequences of external factors but they measure the effect of community actions
on addressing those consequences (e.g., increases in community social cohesion resulting from
greenspace). These indicators include measures relative to intermediate goals and goals directly
related to core values and could be used to monitor outcomes, adaptively manage actions,
demonstrate accountability, and engage the community's interest and support for sustainability.
Type 1:
Indicators of
External Factors
Affecting the
Community
Figure 4.6 Relationship between four types of EPA-identified community sustainability indicators.
By combining these four types of indicators, a community can assess threats, identify priorities, target
actions, demonstrate accountability, monitor results, make informed mid-course corrections, and,
ultimately, measure the impact of the actions in terms of the goals that matter most to the community.
Ultimately, sustainability indicators developed using this approach are expected to help communities
assess whether their actions are helping to sustain and/or move their community toward its core values.
EPA created example frameworks for each of the participating communities as a way to illustrate how to
translate the workshop discussions into a set of sustainability indicators for a community (available upon
request).
W
Type 2:
Indicators of
Consequences
of External
Factors
Intermediate
goals
Core values
*
Type 3:
Indicators of
Possible
Community
Actions
Action Status
Community
response
Type 4:
Indicators of
Outcomes of
Community
Actions
Intermediate
goals
Core values
55
-------
Analysis of how to use workshop outcomes
In addition to providing practical advice on sustainability indicators, the workshops provided
information that could be useful for EPA's research into sustainability indicators and community well-
being.
Develop indices of sustainability
The workshops generated important preliminary insights into the nature and hierarchical structure of
core community values and implications for indices of sustainability. These insights reinforced
information obtained from the previous Regional Sustainable Environmental Science (RESES)
workshops and could help inform further research on sustainability indices. The key findings were as
follows:
• All of the communities participating in the workshops demonstrated an innate capacity for
systems thinking. Without prompting, participants discussed their goals and values in terms of
hierarchies that emphasized inter-relationships among goals and values. Participants consistently
discussed the idea that some goals were "fundamental" or prerequisites for other goals.
Education and Health were cited as goals that need to be achieved in order to attain reasonable
Living Standards, maintain healthy relationships, and achieve other goals. In many cases,
participants stated that all of the goals were important but that community efforts might best be
spent on supporting the fundamental goals. This suggests that in the context of community
decisions and action, values associated with the most fundamental aspects of well-being could be
the highest priorities.
• The sample of communities included different cultural histories, natural settings, and socio-
economic conditions and stressors. Applying the same workshop design to these different
communities revealed the possibility that community values consist of core values that evolve
slowly over time and other values that are prioritized in reaction to stressors or opportunities.
• At any point in time, these more transient values could be perceived as the same or higher
priorities than more stable, core values. In some cases, these more transient priorities may
represent "threshold conditions" that need to be achieved in order to meet other priorities. This
suggests that the landscape of important community values changes over time.
• Participants noted that some important community values were "taken for granted," because they
have been sustained over time. Safety and Security and Social Cohesion were cited as examples
of this. This suggests that values associated with more pressing concerns may cause communities
to lose sight of other important, potentially core community values.
• Practical sustainability indices will need to be adaptable to changes in a way that measures and
emphasizes core values that remain high priorities over time and values associated more
immediate priorities.
The workshops and subsequent analyses also afforded an opportunity to explore the elements of the
HWBI (Smith et al. 2013). They also afforded an opportunity to examine the relative importance values
(RIVs), the factors used to weight different domain scores to derive element scores (e.g., economic well-
being) and an overall HWBI value. Key findings from the workshops are as follows:
• The changing landscape of values discussed above suggests that, from a community perspective,
a set of indices or indicators, rather than aggregated indices, may be more responsive to
community needs. This approach would allow for consistent measures addressing core values
56
-------
and higher and lower prioritization of other indicators as required. The systems thinking
manifested during the workshops suggests that this type of approach would be workable.
The possibility that community values consist of core values that evolve slowly over time and
other values that are prioritized in reaction to stressors or opportunities suggests that RIVs could
also change over time. This suggests the need to periodically update RIVs. It also highlights the
utility of a composite index for measuring well-being versus measuring well-being based on, for
example, a single domain that was initially identified as a priority but may change in priority
over time.
The workshops revealed that different members of a community may have similar priorities with
respect to indicator domains but differences in how they prioritize values associated with the
more detailed indicators presented in the HWBI. Community-wide indicators may be more
representative and stable at a domain level. This suggests that the scope of indicators within each
domain should be adequate to capture differences in the meaning of the domain to different
individuals and care should be taken in defining the scope and combining indicators when
developing aggregate domain indices.
Some participants instinctively made the connection between a healthy natural environment and
human well-being, while other participants struggled with the HWBI definition of "connection to
nature." The latter were more comfortable talking about a healthy natural world as an end in
itself.
Participants consistently suggested that "faith" should be more clearly articulated as possible
core community value. They noted that "rate of congregational adherence" was part of the
cultural fulfillment domain but felt that faith is separate and distinct. The workshop report for
Thibodaux, LA, summarizes a representative discussion.
The workshop discussions, demographic analyses, and community indicators discussion
suggested that the following additional indicators or metrics might be applicable for local
application of the HWBI, if data are available:
o Metrics for physical and mental well-being: anxiety prevalence, physically unhealthy
days, mentally unhealthy days, days with activity limitations due to chronic illness, and
disability status.
o Metrics for healthy lifestyle and behavior: physical activity among adults and youth,
neighborhood walking, residential gardening, and adults and youth eating well.
o Metrics for ability to afford basic necessities: share of income spent on rent and
combined costs of housing and transportation.
o Indicator for cultural fulfillment: connectedness to place (metric: sense of place-based
identity).
o Metrics for responsible engagement in democracy: trust in local government and local
government responsiveness to stakeholders.
The workshop discussions, demographic analyses, and community indicators development
suggested that considerations of frequency of data updates, resolution (e.g., community versus
county), and accessibility will be important in developing indices of well-being to guide
community sustainability decisions and actions.
57
-------
Generalize to the whole community
Engaging a representative group of participants is a critical challenge to be addressed if this type of
workshop is to be effective in identifying the values that are most widely shared by the community. The
methods chapter describes elements of the workshops that were designed to encourage
representativeness, including pre-workshop organizing activities and active facilitation; the results
chapter provides participant-reported assessments of the representativeness of the groups participating in
each workshop; and the discussion chapter provides an analysis of workshop outcomes and socio-
economic data.
The self-reported assessments and the analyses of socio-economic and workshop data provide the
following insights into the within-community generalizability of workshop findings with respect to core
community values:
• The analysis of associations between socio-economic characteristics and workshop outcomes
identified associations that would be logically expected based on a broad community survey.
While the analysis is preliminary and does not account for possible spuriousness, it does suggest
that the workshops can be used to draw preliminary conclusions about broader community
values. Examples of logical associations include:
o Positive linear association between the priority placed on Education (based on mapping
and ranking exercises) and the shares of households in the four participating communities
with children and youth.
o Positive linear association between the priority placed on Health (based on mapping
and/or dot voting), a critical factor affecting household expenses, and the unemployment
rates and percentages of owner-occupied and renter households that spend 35% or more
of their income on housing costs in the four communities.
• The analysis of workshop data revealed no significant bias in terms of higher prioritization of
goals and values that are most closely aligned with the central issues. Observations made during
the workshops suggest that this finding is a result of careful attention to the issue of
representativeness in pre-workshop activities, workshop design and flow, and active facilitation.
Specifically:
o The facilitation team noted that where the central issue had a relatively narrow emphasis
(e.g., recreational environment) participants were more likely to focus on goals that were
closely aligned with the central issue (e.g., Health).
o With active facilitation, the team noted that even where participants started with a narrow
focus, participants were willing and able to broaden their focus and, in general,
understood and saw the value of "building the foundation" (i.e., Part 1 of the workshop)
before focusing on the central issue.
• The comparative analysis of mapping, dot voting, and ranking exercises suggests that the
different types of exercises provide useful and complementary information for generalizing
workshop outcomes to the whole community. The ranking exercise may be a most accurate
expression of core community values. The mapping and dot voting exercises complement the
ranking exercise and provide information needed to develop sustainability indicators. The
progression through the exercises builds understanding and improves the validity of the
workshop outcomes.
58
-------
Regardless of these preliminary findings, the ability to generalize the results of community engagement
workshops, including community-led workshops, to the whole community will be improved by holding
multiple workshops at different times of the day, week, and year and by holding workshops in different
forums. This type of approach will allow for broader community input and will also help distinguish
transient and more stable core community values. In order to support this approach, techniques would
need to be developed for aggregating data across workshops to weigh input appropriately and account
for temporal effects.
Generalize to communities of similar type
One of the goals of the community engagement workshops was to provide information to assess
approaches for classifying communities in ways that link available ecosystem goods and services to core
community values. While it is difficult to draw conclusions from the small number of workshops given
the complexity of the research question, the community workshops highlighted the following
considerations:
• Findings from the workshops affirm the logical presumption that community values are
influenced by a complex mix of history, culture, setting and socio-economic characteristics, and
trends. Workshop observations and data analyses suggest that community values are informed by
historical, cultural, and environmental context; react to socio-economic trends; and involve
systematic relationships among values associated with means versus end goals. This highlights
the importance of a multi-dimensional approach to community classifications that link
characteristics, such as the availability of ecosystem goods and services, and values.
• The strength of the relationship between ecosystem services and community values may be
moderated by important factors, including factors that create inequities in benefits derived from
ecosystem services. For example, comments made by participants in the Pensacola, FL, and Vero
Beach, FL, workshops indicate that ecosystem services provided by the beach and other natural
settings benefit residents unequally. This suggests the importance of accounting for these
moderating factors in a CCS based on ecosystem goods and services.
• Workshop observations and data analyses suggest the importance of considering at least three
different dimensions when considering community values hierarchies and associated
classification systems:
o Historical, cultural, and environmental contexts that define a common frame of reference
for members of the community and could correspond to core values that remain relatively
stable over time.
o Demographic, economic, and other social characteristics that are more likely to change
over time and affect the values hierarchy by emphasizing more transient, but important
community values.
o Foundational values that correspond to goals that communities believe are necessary
prerequisites for achieving and sustaining other outcomes that correspond to important
community values (e.g., Education, Health).
• The above findings suggest the following implications for the structure and substance of a
classification system linking community characteristics and values:
o Certain community socio-economic and other demographic characteristics could have
strong influences on community values and may account for changes in values
59
-------
hierarchies over time. Based on the preliminary analysis, these could include age
distribution, population stability, urban form, socio-economic status, and affordability.
o Participant input during the workshops clearly suggests that historical, cultural, and
environmental contexts provide a common framework for community values. However,
the analysis of workshop outcomes suggests that these contextual factors may be eclipsed
by more immediate and dynamic socio-economic concerns. A multi-level approach that
groups of communities by context (level 1) and includes variables within grouping (level
2) could be effective in establishing a classification system that respects these important
dimensions and accounts for temporal influences on community values.
o While certain contextual and socio-economic characteristics could help explain the most
significant influences on community values, other factors could mediate the relative
priorities of values. These could include factors associated with community-specific
stressors that are viewed as "thresholds" to achieving core community goals. It is likely
that a stronger evidence base would be needed to compile sufficient information for
considering these factors.
o In developing a community classification system, it will be important to consider the
metrics and data sources used to populate variables used to classify communities,
including geographic resolution, data collection frequency, sensitivity, and recognized
validity. The challenge for contextual variables may be defining metrics that sufficiently
distinguish different settings. The challenge for socio-economic variables may be
identifying data sources that capture information at an adequate geographic resolution
and frequency to classify communities in a meaningful way.
• The CCS developed as a part of this study is consistent with these ideas. The context of a
particular community will have to be taken into account in comparisons among CCS groupings.
4.2 Keyword-based analysis of community planning documents
4.2.1 Introduction
Many communities have invested significantly in the compilation of strategic planning documents
intended to help support current and future decision making. These documents usually involve a
thorough examination of the community including solicitation of stakeholder input similar to the
workshops described in Chapter 4.1. It should be possible to describe stakeholder priorities and
community fundamental objectives, and ultimately measures of sustainability success, from an
examination of these documents. Text-based analysis and comparison of documents based on word
frequency is a well-established technique in lexicographic research (Ball 1994, Berber Sardinha 1996).
However, strategic planning documents are not standardized with respect to organization, content, or
scope and these issues must be considered before any meaningful comparison is possible. In this
chapter, a keyword analytical approach was used to examine community strategic planning documents
with the objective of obtaining data on community fundamental objectives similar to that obtained in
Chapter 4.1 from direct engagement. The objective in this chapter is to validate an analytical approach,
use that approach to generate results for a core set of communities, and then compare the outcomes both
between communities and to those reported in Chapter 4.1.
4.2.2 Methods
The keyword analysis of strategic planning documents proceeded in three steps. First, the keyword list
was created and validated through comparison of outcomes from a test set of strategic planning
60
-------
documents. Initial keyword selection was based on the eight domains of the human well-being index
(HWBI, Chapter 3). Second, the keyword list was refined and finalized based on observed outcomes
from this validation analysis. Finally, the completed keyword list was used to analyze a new set of
comprehensive planning documents obtained for 58 communities (city or county level) in the
southeastern United States. The keyword list was initially created based on an examination of five index
community planning documents (Table 4.10). Keywords were extracted from these documents by a
development group (n= 4 people) based on their expert knowledge of the domains in the HWBI. The
development group created the initial keyword list through a line by line examination of a
comprehensive plan for Moss Point, MS (Table 4.10). Word selection was based on previous keyword
development experience of the expert group and a consensus that chosen words clearly reflected a
community value or priority rather than a narrative or an external expert opinion. This initial list was
then culled by collective discussion among the development group to create a test keyword list, which
was then subjected to validation.
Table 4.10 Strategic planning documents used for the validation of the keyword list used in this analysis. See
Chapter 4.2.2 for details. * Document for Moss Point, MS, is at the civic level and was used for initial assembly of
the keyword list but not for validation.
County
State
Source Agency
Comprehensive Plan Link
*Moss Point
MS
Mississippi Gulf
Coast Sustainable
Communities
Initiative
Moss Point Comprehensive Plan; accessed 15 September 2016
Jackson
MS
Jackson Planning
Commission
Jackson County Planner Toolkit; accessed 15 September 2016
Jefferson
FL
Jefferson
Planning
Commission
Jefferson County Comprehensive Plan; accessed 15 September
2016
Terrebonne
LA
Terrebonne
Parish Planning
and Zoning
Terrebonne Parish Comprehensive Plan; accessed 15
September 2016
St. James
LA
St. James County
St. James Parish Comprehensive Plan; accessed 15 September
2016
Wayne
PA
County
Commission
Wayne County Comprehensive Plan; accessed 15 September
2016
*civic plan rather than county
Validation of the test keyword list involved the comparison of manual (i.e., by a person) evaluation of
new strategic planning documents to a keyword-based evaluation of the same documents. The test
keyword list already described was evaluated by comparison of automated and manual reads of six
strategic planning documents not previously used for development of the keyword list (Table 4.10). The
automated read proceeded as described below with the test keyword list. The manual read was
completed by a selected group of five validators each of which read one or two of the five documents.
The validators were asked to extract any statement that they considered a match to a domain of the
HWBI. The validators worked independently and were not given the keyword list prior to completing
their analysis of a document.
Once the manual read was complete the validators met as a group to review and evaluate the results of
the manual read. The results were reviewed for disparity between individual manual reads of the same
document for any domain and questionable matches were discussed and reconsidered for inclusion by
the group. The final results of the two manual reads were combined for each document and compared to
61
-------
the results of the automated read. The manual and automated reads were compared and evaluated for
agreement based on three metrics: relative importance of the eight domains for each document measured
as proportion of total matches per read, number of statements per document found in both read types,
number of statements per document found in manual but not automated read, and vice versa. These
results were examined by the validator group and the keyword list was further amended to improve
agreement between the manual and automated reads. For the purposes of amending the keyword list
(Appendix D), the manual read results were considered most representative of HWBI domain categories.
Once the set of keywords was labelled as final by the validation group, a separate set of community
comprehensive planning documents (Appendix C) was selected and examined using the automated
keyword search combined with the final keyword list. This set of community planning documents was
selected from a systematic search of all counties in the southeastern United States, bordering the Gulf of
Mexico. The documents were found on county websites; if none were available, the county was
excluded. Plans were downloadable in PDF format, either uploaded as images or text. When uploaded as
images, Adobe Pro's Optical Character Recognition was used to change the PDF images into text
format. In each case it was necessary to convert the PDF file (.pdf) into a text file (.txt) before keyword
analysis. Formatting issues and unnecessary information (e.g., historical narrative) were manually
corrected to minimize file conversion errors. Since the objective was to evaluate priorities for future
community actions, narrative elements, such as tables, figures, background history, table of contents,
and appendices, were all manually removed.
The refined documents were analyzed with an original text search algorithm written in the software
package R (R Project; accessed 16 September 2016). The R-script was used to extract statements
containing the selected keywords. The R-script (Appendix E) reformatted each document into one
continuous text string and then separated the string into lines by periods, semi-colons, colons, tabs,
question marks, and exclamation points. The R-script then searched the separated lines individually for
select keywords. Selection was further parameterized by "near" and "exclude" words identified during
the validation process. Phrases were only selected as hits if they included a keyword and a near word. If
one of the exclude words was in the phrase it was not counted as a hit. Each HWBI domain had a unique
set of keywords chosen to characterize that domain and each keyword had multiple near and exclude
words to further refine domain-specific hits. Near and exclude words were initially identified during the
examination of the manual reads, but were further refined during the automatic read. To accomplish that,
all of the current near words were added to the exclude list. By eliminating the words typically returned,
the remaining phrases could be evaluated to create a more thorough compilation of near words to better
encompass the keyword.
Three different metrics were created by the R-script: a count of hits at the indicator level, a count of hits
at the domain level, which comprised hits for all indicators nested under each domain, and a complete
list of every statement categorized as a hit. During the validation phase, the last of these was manually
reviewed to verify the R-script was pulling phrases that fit the intent of the keywords. If a keyword was
pulling hits established as 'false' upon review, the near and exclude words were altered to maximize
positive hits.
Duplicate hits (i.e., same phrase containing multiple keywords and chosen twice for the same domain)
were removed via the R-script for each indicator and domain. The hits were totaled at the domain level
for each separate planning document. To take into account varying lengths of planning documents,
domain hits were normalized per 100 total hits per document. Data from all planning documents were
then combined, summarized, and compared by U.S. state, CCS group (Chapter 2), and differences in
population size, median income, median age, educational attainment, and racial composition of
62
-------
communities. In addition, the combined dataset was analyzed with a principal components analysis (Zar
2010) in order to examine broader patterns involving multiple independent variables.
4.2.3 Results
Keyword list validation - Keyword selection proceeded as described with an initial list of words selected
from a detailed review of the planning document for Moss Point, MS (Table 4.10). This document was
selected as indicative of strategic planning efforts to be targeted for the general analysis and contained a
maximum comprehensive list of focal areas under consideration. Review of this document resulted in a
list of 122 keywords each having between one and 15 near and exclude words respectively. This initial
list was reviewed by the panel and adapted into the test keyword list for validation (Appendix D).
Validation of the keyword list proceeded with analysis of six test planning documents selected randomly
from a working list of publically-available planning documents (Table 4.10). Overall the largest
normalized difference in hits between manual and keyword reads were observed in the Living Standards
domain (Figure 4.7). This domain had a difference exceeding 20% for three documents and exceeding
30% for one (i.e., Wayne County, PA). Maximum normalized difference was below 15% for all other
domains and documents. Living Standards also had the most overall hits among the eight domains with
an average of 38 hits per document. Another notable difference was that for Living Standards the
difference was positive for the manual read meaning the manual results were higher than the keyword
results. With the exception of the Wayne County document, the majority of the other differences were
negative for manual read. The greatest disagreement in results between documents occurred in the
Social Cohesion and Safety and Security domains where three documents were positive for manual read
and three were negative. For a specific document Wayne County showed the most disagreement with the
other documents showing differences in both sign and magnitude for six of the eight domains.
Photo courtesy of U SEP A
63
-------
40
20
a; 30
u
c
(L>
I
=5 io
(L)
S? o
ro
-t-j
c
(L> -10
u
(L>
Q- -20
~ JacksonCo MS
SStJamesPa LA
"~1
fw
S JacksonCo_Neighborhoods_MS
BTerrebonnePa LA
prnr
1 JeffersonCo_FL
] WayneCo_PA
F
X
X
X
n Nsj
X
1: -1 lis!
j a
-30
Connection Cultural Education
To Nature Fulfillment
a
MUa
X
Nl
X
[ill
X
y
Health Leisure Time Living Safety and Social
Standards Security Cohesion
Other
HWBI domains
Figure 4.7 Summary of normalized differences between keyword and manual reads of selected test documents. Zero indicates perfect agreement,
while positive and negative results indicate more or less hits respectively in the keyword analysis for a particular domain. Labels indicate county name
followed by state abbreviation. Jackson County, MS, includes separate analyses at county and neighborhood scale.
64
-------
To further look into the agreement between the human and automatic reads, a line by line examination
was done to match individual phrases between read types (Table 4.11). Agreement is similar when
comparing human and automatic reads to human and human reads when able. In most cases, there were
only one to four matching phrases per domain and document between reads. Of three planning
documents with two human reads each, there were three instances where human to human agreement
was noticeably higher than human to automatic reads. Two of these occurred in the Living Standards
domain while the other one was in Leisure Time. In contrast, there were five instances where human to
automatic reads were noticeably higher than human to human reads. These occurred twice in Health,
then once in each Education, Safety and Security, and Social Cohesion.
Table 4.11 Summary of line by line matches between keyword and manual reads of test documents
organized by HWBI domains. See Chapter 4.2.2 for details.
Keyword
Only
Human
Only
Both
Total
Connection to Nature
26
25
4
61
Cultural Fulfillment
20
28
3
55
Education
39
20
14
94
Health
19
31
13
77
Leisure Time
55
33
10
121
Living Standards
68
196
23
338
Safety and Security
43
63
29
180
Social Cohesion
57
101
15
195
In the Connection to Nature, Cultural Fulfillment, and Education domains, the final keyword list
produced results that were more similar to the human read than the original keyword list. Health, Safety
and Security, and Social Cohesion saw slight deviations, but all reads remained within about two hits of
each other. Only in the Leisure Time and Living Standards domain were the original keyword results
closer to the human reads than the final keyword results. Keyword list finalization involved selection of
the list that generated maximum matches between human and automated reads with a primary focus on
proportional importance of domains and a secondary focus on line matches.
Community comparison of keyword analysis - Keyword-based analysis of 58 planning documents
indicated important differences among domains in a community's stated priorities based on normalized
keyword hits for each HWBI domain (Table 4.12; Figure 4.8). Median proportion of hits across the eight
domains were split into three levels of representation: low, medium, and high for comparison. Cultural
Fulfillment (median value: 3.18) and Health (4.82) were consistently the least mentioned domains. By
contrast, Living Standards (23.61), Safety and Security (17.36), and Leisure Time (15.32) were
mentioned most often. In the middle were Education (8.46), Connection to Nature (9.45), and Social
Cohesion (12.37). Living Standards consistently had the highest median value across all documents
analyzed (Table 4.12).
65
-------
Table 4.12 Summary of normalized keyword hits organized by domains of the human well-being index. See
text for details.
Min
Value
First
Quartile
Median
Value
Third
Quartile
Max
Value
Connection
0.00
5.59
9.45
12.73
19.57
Cultural Fulfillment
0.00
1.66
3.18
6.16
15.00
Education
0.00
6.82
8.46
11.88
28.33
Health
0.67
3.22
4.82
6.49
12.55
Leisure Time
2.86
10.66
15.32
17.78
31.54
Living Standards
12.71
19.29
23.61
29.78
70.29
Safety and Security
0.00
12.83
17.36
22.72
35.51
Social Cohesion
3.43
8.00
12.37
18.69
33.15
80
70
60
50
40
30
20
£
CD
£
3
U
o
T3
0) 50
Q.
CD
CuO
(O
LO
+j
CD
c
u
CD
Q.
I 20
Connection Cultural
Fulfillment
Education
Health
Leisure Time
Living
Standards
Safety and
Security
Social
Cohesion
HWBI Domains
Figure 4.8 Summary of the median across all documents analyzed (58) for normalized hits per domain of
the human well-being index. Bottom and top of boxes indicate 25th and 75th percentile respectively and bottom
and top whiskers indicate the 5th and 95th percentile respectively.
66
-------
Similar patterns among domains are seen when data are organized by state, but differences were evident
between communities grouped by state (Table 4.13). Of the states with more than five documents
analyzed (Florida, Mississippi, and Louisiana), each had Living Standards as the most represented
domains. Florida and Mississippi both had Safety and Security as the second most represented domain,
whereas Louisiana had Social Cohesion ranked second with Safety and Security ranked third near
Education. Of the three states, Florida had the highest score for Safety and Security (19.92) compared to
Mississippi (17.78) and Louisiana (12.44). In contrast, Louisiana had the highest score for Social
Cohesion (19.40), with a notable difference for Florida (10.73) in this domain. An opposite difference
was seen for Leisure Time, with Florida (16.96) and Louisiana (9.87). Louisiana also had the highest
score for Health (6.19) and Education (12.31) in comparison to Florida, which had the highest score for
Connection to Nature (10.65).
Table 4.13 Summary of mean normalized keyword hits for the domains of the human well-being index
organized by U.S. state. See Chapter 4.2.2 for details.
State
FL
MS
LA
AL
GA
TX
Count
36
6
11
2
1
2
Connection
10.65
5.50
8.09
5.63
7.41
4.74
Cultural Fulfillment
3.60
6.14
6.05
2.11
3.17
5.11
Education
8.43
10.31
12.31
5.59
13.23
6.17
Health
4.90
5.53
6.19
4.16
2.65
4.30
Leisure Time
16.96
13.13
9.87
12.05
15.87
15.39
Living Standards
24.81
24.50
25.65
50.14
14.81
20.66
Safety and Security
19.92
17.78
12.44
10.48
17.46
20.40
Social Cohesion
10.73
17.12
19.40
9.84
25.40
23.24
When the data are compared among CCS groups as described in Chapter 2 (Table 4.14), Living
Standards is either the most or second most represented domain in each CCS group, while Cultural
Fulfillment and Health remain the least represented domains similar to the overall results. Yet, among
the CCS groups with at least five documents (Groups 1, 3, 5, and 7) there also were some important
differences. Community classification system Group 1 stands out for Connection to Nature with the
lowest score of 6.57, whereas CCS Groups 3, 5, and 7 each have scores around 11. Community
classification system Group 3 stands out in the Cultural Fulfilment domain with a higher score than
Groups 1, 5, or 7. Education has a larger range among the typologies, with Group 5 (6.69), 7 (7.64), and
3 (8.69) lower than Group 1 (12.33). Wide ranges were present for both Living Standards and Leisure
Time (Table 4.14). Classification system Group 5 was noticeably lower in the Social Cohesion but
higher than the other groups in Safety and Security. Classification system Group 1 is the urban/suburban
category, while Groups 3 and 5 are the most rural. Group 7 is largely suburban but is characterized by a
higher median age and being located almost entirely in the Florida. These differences are largely
congruent with the urban to rural delineation of communities.
67
-------
Table 4.14 Summary of mean normalized keyword hits for the domains of human well-being index
organized by coastal CCS groups. See Table 2.1 for description of community classification system groups.
(Table omits the six city comprehensive planning documents).
Coastal Cluster Group
CG 1
CG 2
CG 3
CG 4
CG 5
CG 6
CG 7
CG 8
Document Count
13
3
9
2
8
1
14
2
Connection
6.57
10.10
11.50
10.36
10.19
16.00
11.12
2.49
Cultural Fulfillment
3.61
5.94
5.89
2.40
3.26
1.00
3.56
4.22
Education
12.33
11.23
8.69
9.46
6.69
13.00
7.64
15.37
Health
5.15
4.99
6.47
3.68
6.12
2.00
4.35
9.52
Leisure Time
10.05
14.37
15.43
14.94
22.32
23.00
15.40
7.79
Living Standards
30.37
19.69
21.88
35.83
21.21
22.00
26.23
29.92
Safety and Security
17.21
13.54
16.48
15.70
20.55
19.00
18.10
15.15
Social Cohesion
14.71
20.14
13.68
7.62
9.66
4.00
13.61
15.53
Of the demographic variables available for comparison across communities, Educational attainment of a
community offers some interesting differences in community priorities. When the communities are
grouped by percent of residents 25 and older with a high school degree or higher or by percent of
residents 25 and older with a bachelor's degree or higher (Table 4.15), similar trends are observed in
number of hits by domain. In each case, a higher score in Cultural Fulfillment is associated with a higher
level of educational attainment. Similarly, counties with the lowest levels of per capita education
completion had fewer keyword hits for Social Cohesion than counties with the two highest levels of
education completion. The reverse of this is true when looking at Education and Living Standards, lower
levels of education had the higher scores for these domains.
Table 4.15 Summary of mean normalized keyword hits for the domains of the human well-being index
organized by proportion of adults citizens with either a high school diploma or a bachelor's degree in 2000
(U.S. Census Data; accessed 14 September 2016). See Chapter 2 for description of community classification
system groups.
High
school
degree or
higher
>80
High
school
degree or
higher
70-80
High
school
degree or
higher
70>
Bachelor's
degree or
higher
>20
Bachelor's
degree or
higher
10-20
Bachelor's
degree or
higher
10>
Connection
8.07
9.85
9.97
9.03
8.61
11.63
Cultural
Fulfillment
5.17
4.11
3.16
5.01
4.39
2.77
Education
8.64
9.22
10.49
8.07
9.99
8.93
Health
4.90
5.73
4.31
5.08
5.45
4.03
Leisure
Time
14.99
14.86
15.20
15.46
13.61
18.92
Living
Standards
24.12
25.76
27.51
23.06
26.17
27.72
Safety and
Security
20.13
15.07
19.61
19.96
16.99
17.42
Social
Cohesion
13.97
15.39
9.77
14.32
14.78
8.59
*Table omits the six city strategic planning documents.
68
-------
Keyword hits also differed among communities by median income, population size, median age, and
demographics, but in a pattern seen in other variables, particularly the CCS groups (Tables 4.16 and
4.17). Scores for Social Cohesion and Cultural Fulfillment were highest for a median income over
$50,000 and the lowest score for median income less than $40,000. Safety and Security and Living
Standards have the opposite pattern, with the highest scores associated with the lowest median income.
Incomes of less than $40,000 also had the highest score for Leisure Time. Education was similar across
all income levels. County population size was broken down into three categories; counties with a
population greater than 150,000, between 50,000 and 150,000, and less than 50,000. Higher populations
were associated with lower scores for Leisure Time. The most variance came from counties between
50,000 and 150,000 people, which had the lowest scores for Connection to Nature and Safety and
Security, as well as the highest score for Living Standards and Education. Median age was split into two
categories (> 40 and < 40) and the older group scored higher for Connection to Nature and Leisure
Time, while the younger group scored higher for Education. Ethnic diversity was measured by relative
proportion of Caucasian, African-American, and Hispanic citizens. As the percentage of African-
Americans increases the scores for both Leisure Time and Connection to Nature decrease. The opposite
is true for Social Cohesion, which increases as the percentage of African-Americans in a county
increases. As the proportion of Hispanics in a community increased the Social Cohesion decreased.
Connection to Nature is noticeably higher when the Caucasian proportion is 70% or greater. Higher
percentages of Caucasians are also associated with a lower score for Cultural Fulfillment and Education
and higher scores for Safety and Security.
Table 4.16 Summary of mean normalized keyword hits for the domains of the human well-being index
organized by median income level, median age, and population size in 2000 (U.S. Census Data: accessed 14
September 2016). See Chapter 2 for description of community classification system groups.
Median
Income
>$50,000
Median
Income
$40,000-
$50,000
Median
Income
$40,000>
Median
Age
>40
Median
Age
40>
Population
>150,000
Population
50,000-
150,000
Population
50,000>
Count
14
23
21
29
29
14
16
28
Connection
9.03
9.05
9.49
11.05
7.35
9.84
6.60
10.37
Cultural
Fulfillment
5.33
4.37
3.58
3.62
5.01
4.43
4.57
4.11
Education
9.87
9.06
9.08
8.07
10.46
9.19
10.32
8.70
Health
4.45
6.19
4.40
4.95
5.29
5.32
5.40
4.86
Leisure
Time
14.84
12.75
17.50
16.15
13.80
12.90
13.24
17.01
Living
Standards
24.21
26.42
25.35
24.50
26.50
25.72
29.75
22.96
Safety and
Security
15.45
17.88
19.63
18.66
17.20
18.99
16.40
18.28
Social
Cohesion
16.81
14.27
10.97
13.00
14.38
13.61
13.72
13.71
69
-------
When variance in all variables, including both community characteristics and keyword-based priorities
by domain, are analyzed together the individual patterns are easily observed (Table 4.17), but there are
interesting pairings among variables suggesting some useful linkages for interpretation. The principal
components analysis demonstrated clear negative relationships between prioritization of Leisure Time
and Education with prioritization of Safety and Security and Connection the Nature. This gradient was
highly associated with a high median age and a high percentage of Caucasian residents, and a high CCS
group number on the Leisure Time end. However, most interesting from this multivariate perspective
was that this gradient was nearly orthogonal with prioritization of Living Standards, as well as mean
income level and population size. The existence of two nearly orthogonal gradients in communities has
some interesting implications for interpretation. Communities in the analysis were well-balanced
between these two gradients in well-being with a slightly higher weight on the former (Figure 4.9).
Table 4.17 Summary of mean normalized keyword hits for the domains of the human well-being index
organized by proportion of community self-reporting in three ethnic groups in 2000 (U.S. Census Data;
accessed 14 September 2016). See Chapter 2 for description.
Caucasian
>80
Caucasian
60-80
Caucasian
60>
African-
American
>30
African-
American
20-30
African-
American
0-10
Hispanic
>20
Hispanic
5-10
Hispanic
0-5
Count
13
26
19
16
20
22
16
13
29
Connection
9.38
10.45
7.36
6.89
9.26
10.83
9.76
10.56
8.28
Cultural
Fulfillment
3.06
4.04
5.55
5.72
4.57
3.06
5.90
3.18
3.96
Education
8.11
9.27
10.06
11.21
9.05
8.04
9.04
7.80
10.04
Health
4.35
6.02
4.43
4.83
5.68
4.83
5.18
4.57
5.34
Leisure
Time
16.42
14.99
13.98
13.16
14.61
16.63
16.13
19.28
12.41
Living
Standards
26.25
23.38
27.89
27.20
24.13
25.50
24.30
23.43
27.09
Safety and
Security
18.66
20.02
14.57
13.22
20.02
19.45
17.21
18.38
18.12
Social
Cohesion
13.76
11.84
16.17
17.75
12.67
11.66
12.47
12.79
14.76
70
-------
-6 -4 -2 0 2 4 6 8
CO
LivingStandards
CO
SafetyandSecurity
0\l
O
-i—i
£Z
o
Cl
E
o
u
o
**.
Healtl
* *
O
o
o
CO
Cl
O
Ed
jcation
LeisureTinie ^
'•j—
Cl
o
Connection- ^
CN
O
(£>
Cultural^j^esion
-0.2
0.1
0.0
0.1
0.2
0.3
Principal component 1
Figure 4.9 Plot summarizing results of a principal components analysis (PC) of normalized hits per domain
of the human well-being index among with the full suite of independent variables considered (See Chapter
4.2.2 for details). Only Principal component 1 (x axis) and Principal component 2 (y axis) are shown. Arrows
(red) indicate direction and size of loadings for independent variables on PCI and PC2 respectively.
4.2.4 Discussion
Keyword analysis of strategic planning documents shows great promise as a contributing method for
clarifying the long-term priorities of stakeholders. Validation results indicated keyword outcomes
generally consistent with a manual read suggesting the approach can be used to interpret planning
documents at least as well as a direct read of the same document. Clarifying community priorities from
document analysis is limited by the scope of the document, as well as the level to which the document
reflects community input rather than the input of elected officials or hired external experts. Yet, these
issues can largely be minimized by appropriate document selection, so the process of reviewing and
selecting documents for analysis should be reported for maximum value of the results. The keyword
results were not biased by document length or text organization, which suggests a wide variety of
documents can be potentially selected for analysis. Exact line-by-line matching of results between
automated and manual reads was less consistent, but it seems the two methods get to the same
interpretation even with some variability in exact phrases aligned with particular domains. The manual
reads were somewhat inconsistent, which created a lot of the discrepancy and suggests an objective
keyword approach should generate more consistent results than interpretation of a document by multiple
individuals.
71
-------
Keyword analysis is an effective technique for analysis of large documents such as the planning
documents considered here, but the technique requires careful analysis. Keyword analysis has been
commonly used to compare documents and corpora for similarity (Berber Sardinha 1996, Peirsman et
al. 2010), to look at word frequency in lexicographic studies (Kilgariff 1996), and use of automated
tools is a well-developed analytical technique (Ball 1994). However these approaches typically involve
analysis of word clusters without pre-defined meaning (Berber Sardinha 1999). The planning document
analysis approach is similar in execution, but is dependent on the use of specific words groups with pre-
assigned meaning. This makes the analysis more dependent on manual vetting of word lists, such as the
validation procedure described earlier (Ball 1994). Therefore, keyword analysis can be very useful, but
its limits must be carefully considered in drawing conclusions.
A key consistency among communities in this analysis was the importance of quality of life metrics to
stakeholder priorities. Across communities and community types the consistently dominant domains, in
terms of total number of keyword hits, were Living Standards followed by Safety and Security followed
closely by Social Cohesion and Leisure Time. An interest in quality of life seems to be a common
community attribute, which is not surprising. The consistent low scores for either Connection to Nature
and Health were surprising, but suggest these are not community-level priorities (i.e., not highlighted in
comprehensive plans) but may be important at a different scale (e.g., personal/family). For instance,
even in cases where an action may directly benefit human health (e.g., investment in hospitals) the
community-scale priority for the action may not be directly tied to health, but rather to ancillary benefits
more aligned with community-scale priorities such as job creation, reductions in burden on public
services, or community reputation. These differences can be important to setting measures of success at
the appropriate scale. It is also important to understand if these results differ among community types
(Bagstad and Shammin 2012).
The dominant delineations for stakeholder priorities at the community level were between states and
CCS groups (Chapter 2). The state comparison made here is limited to three states in the coastal Gulf of
Mexico region (LA, MS, and FL), but this allows for a useful comparison of state differences to
differences in other categories. For CCS, four groups could be meaningfully compared. States differed
most for Safety and Security and Social Cohesion, while CCS groups differed most in Living Standards
and Leisure Time. The less commonly mentioned domains such as Connection to Nature were more
important in specific categories such as median age and ethnic composition of the community. The
broader categories, such as CCS, are not independent from these specific categories (e.g., demographics)
as these data were also used as a part of the CCS score calculation (Chapter 2). However a lack of
pattern at the CCS level suggests these issues are more individualistic in nature and may not drive
community decision making other than to crystalize individual interests.
The value of understanding these differences among groups is to identify the domains of human well-
being for which the CCS or geographic delineations are the most informative. These most informative
differences lie on a gradient from an emphasis on Safety and Security and Living Standards on one end
to an emphasis on Leisure Time and Social Cohesion on the other end. This generalization is supported
by both the categorical results, as well as the multivariate analysis of all domains together. This gradient
is also consistent with an urban to rural gradient in that it is directly related to population size, and
demographics as 'ruralness' tends to be related to an increased emphasis on social connectivity (Bagstad
and Shammin 2012, Smith and Clay 2010). As communities become more urban, more diverse, or less
dependent on local natural resources, they also tend to have a higher median income and a higher overall
education level, which is consistent with, but not dependent on increasing urbanization. They also seem
to prioritize Safety and Security, Living Standards, and Connection to Nature; and reduce priorities for
Social Cohesion and Education. The inverse relationship between educational attainment in a
72
-------
community and their interest in Education was surprising but fairly consistent, which suggests it is a real
trend among groups.
In absolute terms the most informative delineation of keyword data at the community scale is for CCS
groups followed by state differences, but other delineations become more important at smaller scales
within the community. Domains such as Connection to Nature and Education do not parse out very well
at the community scale, as indicated by the lack of difference among communities for these domains,
and the lack of information about them contained in categories such as CCS and geography.
Nonetheless, they can be quite important in driving individual priorities and so have a collective
influence at the community level not well captured by review of community planning documents. As
such, it is not advised that any conclusions can be drawn about community priorities for these domains
with a keyword-based method. Better data may be obtained if more specific documents were used for
the analysis and this is a topic for future study. These findings strongly suggest that keyword analysis
combined with a CCS based comparison can be very informative regarding differences in the relative
importance of community-scale priorities such as Social Cohesion, Living Standards, Leisure Time, and
Safety and Security. Beyond the specifics, it is evident that communities differ in how they rank and
prioritize the domains of human well-being and these differences are predictable based on community
type. This indicates the value of community delineations for informing the decision process. However, it
also indicates that measures of success can only be partially generalized and the very definition of
human well-being may differ among community types as has been pointed out in the past (Bagstad and
Shammin 2012, Moller et al. 2015). Such differences must be kept in mind when comparing the
objective well-being across communities, particularly along the urban to rural gradient. Therefore, use of
this technique in the future should focus on improving the understanding of how community type may
inform differences in the importance of the domains of human well-being that can be used to both
develop and assess decision options at the community level.
4.3 Engagement conclusions
In this section, two methods for obtaining stakeholder input on community level priorities were
explored. The human well-being index (HWBI) was used as a framework for engagement in each case,
but the source of information was very different. In Chapter 4.1, a workshop approach is described
based on structured decision making (Structured Decision Making; accessed 16 September 2016), while
in Chapter 4.2 an automated analysis of strategic planning documents is described based on keyword
counting method. Both methods have advantages and disadvantages for identifying stakeholder
priorities, but more importantly they may be highly complementary methods that should be considered
for paired application.
Key differences were observed in the outcomes of these two methods when they were applied to the
same communities across community type (Figure 4.10). The workshop method generated more diverse
findings that nonetheless consistently reported high importance in the domains of Education and Social
Cohesion. In contrast, the keyword method was always dominated by the Living Standards domain,
which is the primary economic domain of the HWBI. In terms of meaning, the keyword results are
based on strategic planning, which is predictably action-focused and heavily weighted to economic
aspects of a community's well-being. In contrast, workshop results show a broader influence and this is
likely the result of facilitation and the separation of community priorities from a particular action
(Chapter 4.1). However, if one removes the Living Standards domain from the keyword results, the
relative importance of the remaining domains is highly congruent with workshop results (Figure 4.10).
Therefore, the findings of the keyword analysis can be thought of as hierarchical with the secondary
outcomes being more similar to workshop outcomes. There is, thus, strong support for the
complementarity of the two methods.
73
-------
to
CD
£ 35%
-5. 30%
to
25%
^ 20%
S 15%
go 10%
¦2 5%
g 0%
&_
CD
o_
Thibodaux, LA
1 a §
J? ^
S ^ *
^ ¦*
<0^
r>^
va
J*
Pensacola, FL
1
./ / .
w6 ^ ^ <-°
J? J
<** * # 25%
•*= 20%
•s 15%
i/i 10%
(D
uo 5%
ro
£ 0%
(D
u
1_
(D
CL
yj
&
c^°
/
^ an<
u
&
Vero Beach, FL
alljk
A
-x<^
&
4<2r
*F
HWBI domains
s /
<.\> O-"
o< • .v
/y _<>
<* S°
Figure 4.10. Summary comparison of combined results from workshops (Chapter 4.1) and keyword analysis (Chapter 4.2) in four focal
communities. Community name is given in each panel. Three bars are results from an individual ranking exercise during the workshop (hashed; WS rank),
results from a group voting exercise during the workshop (solid; WS dot vote), and results of the relative proportion of hits from the keyword analysis
(diamond; Keyword). Details of these data can be found in the respective chapters.
74
-------
This complementarity is also highlighted in the advantages and disadvantages of each engagement
method. The keyword method is based on existing and vetted information obtained from extensive
discussion and feedback of community stakeholders. Often the values reported in these planning
documents are the results of months of stakeholder engagement occurring in multiple forms (e.g.,
workshop, mail survey, interviews). However, the information is usually obtained in a very constrained
way based on particular issues of current importance to a community. The information is also filtered
through community leaders and sometimes outside experts, and may be packaged to meet pre-conceived
objectives. Further, the findings are dependent on the keyword list which must be vetted and tested prior
to use. In contrast, the workshop method is well planned and facilitated to obtain a broad representative
outcome reflecting community priorities independent of any particularly issue. The information is
therefore more comprehensive than what may be found in planning documents. However, the group of
people contributing to a workshop is far smaller than for the keyword analysis and may not represent the
entire community. In balance, a consistent result across the two approaches provides good support for
the complementary nature of the data and the value of applying both methods simultaneously to identify
community priorities.
The priorities were more consistent across communities than across community types. For both
analytical methods, community type was most informative about the relative importance of low scoring
domains of the ITWBI such as Connection to Nature and Cultural Fulfillment. This is important
information for scoring the HWBI and will be used to explore relative weighting within the HWBI, but
the dominance of Living Standards, Safety and Security, Education, and Social Cohesion was consistent
in both community type examined and so seems robust to categorization. Community-specific
deviations were more evident, such as the dominance of the Health domain in Vero Beach, FL, and the
Living Standards domain in Opelousas, LA. However, these community-level differences are to be
expected and the overarching consistency of multiple domains across communities suggests important
common themes that should be explored for their value in informing and measuring the success of
community level decision support.
Photo courtesy of USEPA
75
-------
4.4 Literature cited
Bagstad, K.J. and M.R. Shammin. 2012. Can the genuine progress indicator better inform sustainable
regional progress? A case study for Northeast Ohio. Ecological Indicators 18:330-341.
Ball, C.N. 1994. Automated text analysis: Cautionary tales. Literary and Linguistic Computing 9:295-
302.
Berber Sardinha, A.P. 1996. Review of WordSmith tools. Computers and Text 12:19-21.
Berber Sardinha, A.P. 1999. Using Keywords in Text Analysis: Practical Aspects. DIRECT papers,
working paper 42 ISSN 1413-442x. CEPRIL, PUC-SP, Brazil, and AELSU, Liverpool
University, England.
Carriger, J.F., W.S. Fisher, T.B. Stockton Jr., and P.E. Sturm. 2013. Advancing the Guanica Bay (Puerto
Rico) watershed management plan. Coastal Management 41:19-38.
Fang, K., R. Heijungs, and G.R.d. Snoo. 2014. Theoretical exploration for the combination of the
ecological, energy, carbon, and water footprints: Overview of a footprint family. Ecological
Indicators 36:508-518.
Gregory, R.S. and R.L. Keeney. 2002. Making smarter environmental management decisions. Journal of
the American Water Resources Association 38:1601-1612.
Holden, M. 2013. Sustainability indicator systems within urban governance: Usability analysis of
sustainability indicator systems as boundary objects. Ecological Indicators 32:89-96.
Kilgariff, A. 1996. Using Word Frequency Lists to Measure Corpus Homogeneity and Similarity
Between Corpora. COLING workshop on very large corpora. Information Technology Research
Institution, University of Brighton, UK.
Moller, V., B. Roberts, and D. Zani. 2015. The personal wellbeing index in the South African IsiXhosa
translation: A qualitative focus group study. Social Indicators Research 124:835-862.
Peirsman, Y., D. Geeraerts, and D. Speelman. 2010. The automatic identification of lexical variation
between language varieties. Natural Language Engineering 16:469-491.
Smith, C.L. and P.M. Clay. 2010. Measuring subjective and objective well-being: Analyses from five
marine commercial fisheries. Human Organization 69:158-168.
Smith, L.M., H.M. Smith, J.L. Case, and L. Harwell. 2012. Indicators and Methods for Constructing a
U.S. Human Well-Being Index for Ecosystem Services Research. U.S. Environmental Protection
Agency, Washington, DC, EPA/600/R-12/023.
Smith, L.M., J.L. Case, H.M. Smith, L.C. Harwell, and J.K. Summers. 2013. Relating ecosystem services
to domains of human well-being: Foundation for a US index. Ecological Indicators 28:79-90.
Vackar, D., B. tenBrink, J. Loh, J.E.M. Baillie, and B. Reyers. 2012. Review of multispecies indices for
monitoring human impacts on biodiversity. Ecological Indicators 17:58-67.
Zar, J.H. 2010. Biostatistical Analysis, 5th edition. Prentice Hall, New Jersey.
76
-------
5 Ecosystem goods and services
5.1 Introduction
A critical element of community decision making is the actual and perceived value of goods and
services being collected from the surrounding ecosystem. Available ecosystem goods and services
(EGS) differ greatly in form and visibility to community stakeholders (Millennium 2005). For instance,
a community may be highly aware of an extractive resource such as a fishery, as it may be an economic
driver for the community. Yet, less visible services such as healthy swimmable waterways, or
viewscapes and their contribution to community identity, may be just as important in driving community
priorities. All ecosystem goods and services contribute in some way to community well-being, but those
EGS most common across coastal communities that can be most useful for comparing communities and
looking for common themes are of greatest interest. Therefore, this chapter focuses on a comparison of
four focal communities (see Chapter 4) with respect to the availability of four critical EGS: flood
protection, usable air, usable water, and stable climate. These four critical EGS are highly valuable to all
coastal communities (Russell et al. 2013), but also highly dependent on local conditions with respect to
land cover and quality (Barbier et al. 2011). First, the value and distribution of these core EGS in each
community is characterized. Then, a across community comparison was conducted to look for patterns
among community types.
5.2 Methods
Estimates of service delivery and value were collected in the four service categories for four selected
counties in the coastal Gulf of Mexico region. The service categories were: flood protection, usable air,
usable water, and stable climate. The four counties were used in the comparison and described in
Chapter 1. Production rate and value of selected EGS were estimated based on the composition of land
use, soils, impervious surfaces and canopy cover within each county and value based on decreased
health care costs, social coast of carbon and replacement values (Russell et al. 2013). Land use was
based on National Land Cover Dataset (NLCD, Homer et al. 2015) maps for each county. Rates and
values for individual EGS were either obtained from the literature or calculated for each land use type
based on secondary data. The land use maps were downloaded directly from the NLCD website and
imported into ArcGIS (ESRI Version 10.1) as rasters. Data on impervious surface cover, canopy cover,
and soil classification were also obtained as secondary data to NLCD (Table 5.1). In order to generate
EGS values, each of the NLCD classes were selected by attribute and then exported as an individual
map layer. This was necessary because the files from the NLCD websites are rasters, which prohibits
them from being spatially joined. To match all data layers, canopy cover, soil type, and impervious
surface map grids were extracted using NLCD land use masks created for each county.
77
-------
Table 5.1 Summary of external data used in calculation of ecosystem goods and service delivery. Data were
all downloaded separately for the four focal counties. See text for details.
Data
Source
URL
NLCD Land Use
NLCD 2011
Land Cover
National Land Cover Database 2011; accessed 15 September
2016
Denitrification
Rates
See Table 5.5
N/A
Carbon Burial
See Table 5.6
N/A
Impervious Land
Cover
NLCD 2011
Percent
Developed
Imperviousness
National Land Cover Database 2011; accessed 15 September
2016
Canopy Cover
NLCD 2011
USFS Tree
Canopy
Cartographic
National Land Cover Database 2011; accessed 15 September
2016
Soil
Classification
Soil Survey
Geographic
Database
USDA Web Soil Survey Database; accessed 15 September 2016
Curve Number
Zhang et al.
2011
N/A
5.2.1 Usable water and stable climate
The services usable water and stable climate were based on the effective sequestration or removal of
excess nitrogen and carbon respectively as a function of land use category (Russell et al. 2013). The
denitrification and carbon burial rates were average values for each NLCD category derived from a
comprehensive literature review (Appendix F). Denitrification and carbon burial values selected for each
of the NLCD categories (Appendix G) came from peer-reviewed literature sources, predominantly work
located in the southeastern United States / Gulf of Mexico region. The rate values included in each
average were individually selected based on relevance to the county, making the table county-specific
wherever possible. The selected rates within all NLCD categories, except developed land, were then
averaged as the overall rate. Coverage of NLCD categories for each county are given in Appendix G.
Denitrification and carbon burial rates for the developed land classes required further calculations to
account for the extent of impervious surface within each land use category. Denitrification and carbon
burial rates for the developed land use categories were based on the literature-based average for open
land (e.g., non-impervious) adjusted for the amount of impervious surface estimated to be present in a
particular developed land-use category in a particular county. The impervious cover maps were
downloaded directly from the NLCD website (Xian et al. 2011). To assign an impervious cover
percentage to each pixel, the pixels were extracted by a mask for each of the NLCD classes created
above. The impervious cover percentages for that set of pixels were averaged and recorded for each
NLCD class per county and compiled into a master table. The denitrification and carbon burial rates for
developed open space were first calculated as described above based on literature values. This average
was then adjusted according to the following equation:
78
-------
(1 - % Impervious) * (Mean denitrification or carbon burial rate for open space)
Which results in either a denitrification or carbon burial rate value per square meter. Since the percent
impervious surface varies between each of the developed land classes, the rates are NLCD class and
county specific. The usable water value is based on the NLCD category specific denitrification rates and
was calculated using the following equation:
0.018 ($ / g N) * Denitrification Rate * (10,000 m2 / ha)
The denitrification value term based on abatement costs of reducing nitrogen from point sources
estimated at $8.16 / lb, which equals $18 / kg (Birch et al. 2011). The stable climate value is a
parameterization of the mean carbon burial rate and was calculated using the following equation:
1.3542 * 10"4 ($ / g C)* Carbon Burial Rate * (10,000 m2 / ha)
The carbon burial value term was generated from the value of carbon removal: $37 / ton of carbon
dioxide (Boscolo et al. 1998). Since the molecular mass of carbon dioxide (44.0095 g / mol) is 3.66
times as massive as carbon (12.0107 g / mol), the cost per ton of carbon is $37 * 3.66 = $135.42 / ton C.
5.2.2 Usable air
The usable air value is the representation of the decrease in health care costs resulting from removal of
five common pollutants: carbon monoxide (CO), ozone (O3), particulate matter (PM10), sulfur dioxide
(SO2), and nitrogen dioxide (NO2) (Murray et al. 1994). The usable air service is based on the removal
value per average canopy cover within a NLCD land use class. The canopy cover percentages were
determined by extracting a mask of each of the NLCD land use classes from the NLCD canopy cover
map downloaded from the NLCD website (Homer et al. 2015). The canopy cover percentages for each
set of pixels were averaged and recorded for each NLCD class per county and compiled into a master
table.
For each pixel, the removal rate for the particular pollutant was multiplied by the percent canopy cover,
the area, and the estimated value of removal. These values were added together and consolidated in one
equation as seen above to estimate usable air.
CO: (0.5 g / m2 / yr) * (% Canopy Cover) * (1 ton / 1,000,000 g) * (Area in m2) * (959 $ / ton) = $/yr
O3: (5.8 g / m2 / yr) * (% Canopy Cover) * (1 ton / 1,000,000 g) * (Area in m2) * (6,752 $ / ton) = $/yr
SO2: (2.4 g / m2 / yr) * (% Canopy Cover) * (1 ton / 1,000,000 g) * (Area in m2) * (1,653 $ / ton) = $/yr
PM10: (4.5 g / m2 / yr) * (% Canopy Cover) * (1 ton / 1,000,000 g) * (Area in m2) * (4,508 $ / ton) =
$/yr
NO2: (1.1 g / m2 / yr) * (% Canopy Cover) * (1 ton / 1,000,000 g) * (Area in m2) * (6,752 $ / ton) = $/yr
Which combine into the following overall formula:
Total: [(0.000713215 $ / m2 / yr) * (% Canopy Cover) * (Area in m2)] * (10,000 m2 / 1 ha)
79
-------
5.2.3 Flood protection
The flood protection service is the value of the grey infrastructure needed to be built to handle the
amount of water from the average two year period storm event if the natural system were not present
(Wang et al. 2013). This is based on the curve number, a function of soil type, for each of the NLCD
land classes.
The soil type maps were based on the soil survey area (soil survey geographic database = SSURGO)
data downloaded from the USDA website (U.S. Department of Agriculture 2015). Soils are classified
into different hydrological groups, distinguished by letters A through D, based on soil texture. In
Microsoft Access, the file was opened and the tables were imported into the soil's tabular folder. The
component table was exported as a text file with a .csv extension. Once the component table was added
to a GIS platform along with the original soil shapefile, the two were joined based on their map unit key
(MUKEY) fields. All fields except the hydro groups and map unit keys were deleted. A new field,
Max Type N, was added to the attribute table. Then, selecting by attribute from the hydro group field,
all attributes with A or A/D type soil were assigned a value of 1 in Max_Type_N. The same applied for
the other hydro groups: B or B/D, C or C/D, and D were assigned a value of 2, 3, or 4, respectively. The
map was then exported as a new shapefile and the symbology was changed based on Max_Type_N
value.
Curve number (CN) is an index that represents an area's ability to hold water. That is, after a rain event
has begun and just before runoff starts to occur, how much water has entered the system. The lower a
CN is, the lower its runoff potential. For example, a land use type with a CN of 30 has a very high water
retention rate and a low runoff potential, whereas a land use type with a CN of 100 has a very low water
retention rate and a high runoff potential. The curve number maps were based on the Max Type N
values from the above Soil Type attribute tables and the NLCD class. According to Zhang et al. (2011)
there are four potential curve numbers for each NLCD class based on the four different Max Type N
values. To align these, a new field was added to the new Soil Type shapefile created above. It was
populated by the following equation:
(100* Max_Type_N) + 2 digit NLCD code
This merges the two fields together to create a three digit code. The Zhang et al. (2011) table was
reorganized in an identical fashion and the two tables were joined by code, assigning a curve number to
each pixel which were then totaled for each county (Appendix I).
The Flood Protection Value was calculated using the curve number as described above and the
following equation:
[(0.05 * [(25,400/CN) - 254)] / (1,000 mm/ 1 m)] * 70.629265 ($/mm/m2) / (30 yr) * (10,000 m2/ 1 ha)
This was generated from multiplying the depth of the water retained in the area that was affected and the
value of water retention per cubic meter. This was achieved by converting the water retention value
from $2/ft3 (Wang et al. 2013) to $ / m3 by multiplying by 35.3147, totaling $70.629265 / m3. The depth
of the water retained in mm for each feature is (0.05 * (25,400 mm / Curve Number) - 254 mm). 30 year
is based on the design lifespan of the grey infrastructure (Wang et al. 2013).
80
-------
5.3 Results
The four counties examined differed in both the amount and distribution of NLCD land cover types
(Figure 5.1). The three largest land cover types in Escambia County, FL (Pensacola), are Evergreen
Forest (22.06%), Woody Wetlands (18.55%), and Shrub / Scrub (13.93%). Of the remaining twelve land
cover types, seven of them are under 3% of the total land coverage individually. In Indian River County,
FL (Vero Beach), the two most prevalent land cover types are Cultivated Crops (28.52%) and Woody
Wetlands (27.02%). Similar to Escambia County, Indian River County has eight land cover types each
under 3% of total area. Lafourche Parish, LA (Thibodaux), is even less diverse. The top three areal land
cover types are Emergent Herbaceous Wetlands (42.57%), Woody Wetlands (22.29%), and Open Water
(15.98%)). There are nine land cover types that are each under 1% of the total land coverage. The three
most dominate land cover types for St. Landry Parish, LA (Opelousas), are Cultivated Crops (39.49%),
Woody Wetlands (32.07%) and Pasture / Hay (14.72%) (Table 5.2). Of the remaining land cover types,
ten account for less than 1% individually of the total land area. It is important to note that the
percentages are proportional values based on each county's total area. Escambia County and Indian
River County are roughly half the area of Lafourche Parish and St. Landry Parish, respectively. For
example, although woody wetlands was the second most dominate land cover type in both Escambia
County and Lafourche Parish and differed by less than 4%, the areal coverage in Lafourche Parish was
twice as much as in Escambia County. Furthermore, as a generality, all four counties are dominated by
undeveloped or cultivated land with a consistently low percentage of development. This is important
when considering the EGS on a per capita basis. There are also differences in the percentage of
impervious land cover and areal canopy cover between the four counties. This is based on the amount of
coverage of each for the pixels classified in each NLCD category. For example, impervious land cover
percentage for Open Space of Escambia County is 8%, while it is 13% for Lafourche Parish (Table 5.3).
Further highlighting differences, the percentage of areal canopy cover for woody wetlands in Indian
River County is 57% and 90% in St. Landry Parish (Table 5.4). These differences have an effect on the
values of usable water and air for each of the counties.
81
-------
Condensed NLCD Classes
| Water
| Developed
~ Barren
| Forest
m Open Vegetation
~ Agriculture
H Wetlands
Figure 5.1 Maps showing National Land Cover Dataset (NLCD) coverage (See Table 5.2) for: (A) Escambia
County, FL; (B) Indian River County, FL; (C) Lafourche Parish, LA; and (D) St. Landry Parish, LA. The
NLCD classes were condensed for clarity to a single color for the legend and pie charts, but not for the maps.
Water is Open Water; Developed is Open Space, Low, Medium, and High Intensity Developments; Barren is
Barren Land; Forest is Deciduous, Evergreen, and Mixed Forests; Open Vegetation is Shrub / Scrub and
Herbaceous; Agriculture is Hay / Pasture and Cultivated Crops; Wetlands is Woody and Emergent Herbaceous
Wetlands.
82
-------
Table 5.2 Total area and percentage of areal coverage of NLCD land cover categories for four counties. Water is Open Water; Op Spc is Open
Space; L Dev is Low Intensity Development; M Dev is Medium Intensity Development; H Dev is High Intensity Development; Barrn is Barren; De
For is Deciduous Forest; Ev For is Evergreen Forest; Mi For is Mixed Forest; Shrub is Shrub/Scrub; Grass is Grassland/Herbaceous; Pastur is
Pasture/Hay; Crops is Cultivated Crops; W Wet is Woody Wetlands; E Wet is Emergent Herbaceous Wetlands.
NLCD
Escambia
Area(ha)
Escambia
Percentage
(%)
Indian River
Area(ha)
Indian River
Percentage
(%)
Lafourche
Area(ha)
Lafourche
Percentage
(%)
St. Landry
Area(ha)
St. Landry
Percentage
(%)
Water
2,122.74
1.23
7,112.79
5.32
48,129.84
15.98
3,926.97
1.61
Op Spc
20,562.48
11.88
11,869.47
8.88
2,047.05
0.68
6,698.16
2.75
L Dev
11,987.10
6.93
7,359.39
5.51
8,693.01
2.89
9,338.13
3.84
M Dev
4,974.93
2.87
2,460.51
1.84
1,179.99
0.39
527.58
0.22
H Dev
1,997.01
1.15
487.44
0.36
846.72
0.28
365.04
0.15
Barrn
2,452.05
1.42
536.94
0.40
784.44
0.26
38.25
0.02
De For
591.21
0.34
20.52
0.02
159.57
0.05
2,805.03
1.15
Ev For
3,8175.39
22.06
2,474.73
1.85
34.92
0.01
550.71
0.23
Mi For
2,471.85
1.43
137.88
0.10
124.65
0.04
1,460.61
0.60
Shrub
24,111.99
13.93
2,892.51
2.16
856.62
0.28
4,578.39
1.88
Grass
7,740.27
4.47
10,52.55
0.79
454.95
0.15
1,677.96
0.69
Pastur
6,766.29
3.91
10,948.14
8.19
13,532.67
4.49
35,805.60
14.72
Crops
14,955.12
8.64
38,175.93
28.56
28,982.79
9.62
96,048.36
39.49
W Wet
32,105.52
18.55
36,109.98
27.02
67,128.03
22.29
77,996.16
32.07
E Wet
2,071.26
1.20
12,016.71
8.99
12,8211.12
42.57
1,396.17
0.57
Total
173,085.2
100
133,655.49
100
301,166.4
100
243,213.1
100
-------
Table 5.3 Percentage of impervious land cover in each of four counties. Data are based on published
values from NLCD 2011 Percent Developed Imperviousness data for each developed land cover category.
This is used in the calculation of denitrification and carbon burial rates for developed land classes.
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
NLCD
Percentage (%)
Percentage (%)
Percentage (%)
Percentage (%)
Open Space
8
8
13
12
Low Intensity
33
32
33
30
Development
Med Intensity
61
59
61
62
Development
High Intensity
87
86
85
85
Development
Table 5.4 Percentage of areal canopy cover by land cover category for four counties. Data are
calculations based on published values from NLCD 2011 USFS Tree Canopy cartographic data for each
developed land cover category. This is a metric for the valuation of air quality.
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
NLCD
Percentage (%)
Percentage (%)
Percentage (%)
Percentage (%)
Open Water
0
0
0
0
Open Space
38
24
10
26
Low Intensity
16
14
7
12
Development
Med Intensity
4
5
1
2
Development
High Intensity
1
1
0
1
Development
Barren
6
5
1
13
Deciduous Forest
86
75
56
84
Evergreen Forest
82
55
35
81
Mixed Forest
82
34
64
82
Shrub/Scrub
60
26
23
58
Grassland/Herbaceous
25
11
11
19
Pasture/Hay
14
5
8
11
Cultivated Crops
5
9
4
7
Woody Wetlands
83
57
77
90
Emergent Herbaceous
30
14
4
46
Wetlands
84
-------
Table 5.5 Denitrification rates (g N/m2/yr) by NLCD category for each county. Calculations are based
on published rates for each land cover category. This is a metric for the valuation of water quality.
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Open Water
8.29
8.29
8.29
8.29
Open Space
0.71
0.71
0.67
0.67
Low Intensity
Development
0.51
0.52
0.52
0.54
Med Intensity
Development
0.3
0.31
0.3
0.29
High Intensity
Development
0.1
0.12
0.12
0.12
Barren
1.05
1.05
1.05
1.05
Deciduous Forest
0.22
0.22
0.22
0.22
Evergreen Forest
0.09
0.09
0.09
0.09
Mixed Forest
0.14
0.14
0.14
0.14
Shrub/Scrub
0.92
0.92
0.92
0.92
Grassland/Herbaceous
0.19
0.19
0.19
0.19
Pasture/Hay
4.31
4.31
4.31
4.31
Cultivated Crops
10.23
10.23
10.23
10.23
Woody Wetlands
17.21
17.21
17.21
17.21
Emergent Herbaceous
Wetlands
12.92
12.92
12.92
12.92
Table 5.6 Carbon burial rates (g C/m2/yr) for each NLCD category in four counties. Data are based
on published rates for each land cover category. This is a metric for the valuation of a stable climate.
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
NLCD
Open Water
103.25
103.25
210
210
Open Space
91.75
91.62
86.93
87.2
Low Intensity
66.55
67.75
66.96
69.8
Development
Med Intensity
39.15
40.7
38.52
38.08
Development
High Intensity
12.67
14.96
14.94
15.04
Development
Barren
0
0
0
0
Deciduous Forest
7.97
7.97
7.97
7.97
Evergreen Forest
47.14
47.14
47.14
47.14
Mixed Forest
27.56
27.56
27.56
27.56
Shrub/Scrub
0
0
0
0
Grassland/Herbaceous
30.11
30.11
30.11
30.11
Pasture/Hay
48.65
48.65
48.65
48.65
85
-------
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Cultivated Crops
43.48
43.48
43.48
43.48
Woody Wetlands
171.53
171.53
171.53
171.53
Emergent Herbaceous
Wetlands
187.14
187.14
187.14
187.14
Spatial distribution of land cover classes varied between counties particularly in the placement
and concentration of developed classes (Figure 5.1). Escambia County, FL, is densely populated
in the southeastern portion of the county with scattered agriculture and pasture in the north.
Indian River County, FL's developed areas are isolated on its eastern coast. The vertical middle
of the county is predominantly cropland and the western part of the state is a mixture of open
water, pastureland, and wetlands. Lafourche Parish, LA, has a narrow band of development
adjacent to open water and surrounded by agriculture and pastureland running along the center of
the county. Surrounding this is wetlands and open water. St. Landry Parish, LA, has large
portions of wetlands in its northern and eastern sections. Its central and western landscape is
mostly agriculture and pastureland. There is a small pocket of development in its westernmost
section and another in its center.
Separating the values of ecosystem goods and services further by NLCD type, the only
differences in the value of maintaining water quality are in the development classes (Figure 5.6;
Table 5.7). This is because the calculation considered differences in canopy cover and
impervious surfaces among counties in estimating denitrification rates (Figures 5.2 and 5.3;
Tables 5.3 and 5.4). The largest per hectare values for usable air in NLCD were for the forest
classes, shrub / scrub, and woody wetlands (Figure 5.7; Table 5.8). Similar to the usable water
values, the values for maintaining stable climate only varied in open water and the development
classes (Figure 5.8; Table 5.9). The variation in open water values was the result of separation of
carbon burial rates based on ecosystem location. For example, the two coastal communities
factored in open ocean carbon burial rates to the average, while the two inland communities
included lake rates. For the development classes, as with maintaining water quality values, the
variation is a reflection in differences in canopy cover and impervious surfaces. Across the
counties, flood protection was highest in the forest classes. Open space, shrub / scrub, grassland,
and pasture NLCD classes provided mid-level flood protection. Open water, woody wetlands,
and emergent herbaceous wetlands provided no flood protection (Figure 5.9; Table 5.10). This is
because those ecosystems are already flooded and cannot absorb any more water. This is
opposite of the aforementioned land cover types that usually are not flooded and can therefore
retain a larger volume of rain water, thus preventing runoff.
86
-------
Impervious %
Low : 0
High : 100
Figure 5.2 Impervious land cover as reported by the NLCD 2011 Percent Developed
Imperviousness data (See Table 5.3) for: (A) Escambia County, FL; (B) Indian River County, FL;
(C) Lafourche Parish, LA; and (D) St. Landry Parish, LA. This is used in the calculation of
denitrification and carbon burial rates for developed land classes.
87
-------
Canopy Cover %
High : 100
Low: 0
Figure 5.3 Canopy cover as reported by the NLCD 2011 USGS Tree Canopy cartographic data (See
Table 5.4) for: (A) Escambia County, FL; (B) Indian River County, FL; (C) Lafourche Parish, LA;
and (D) St. Landry Parish, LA. This is a metric for the valuation of air quality.
88
-------
DenJtrlfication Rates
(g N / m21 yr)
I 0.09
Figure 5.4 Denitrification rates averaged from literature review (See Table 5.5) for each NLCD
category for: (A) Escambia County, FL; (B) Indian River County, FL; (C) Lafourche Parish, LA;
and (D) St. Landry Parish, LA. This is a metric for the valuation of water quality.
89
-------
Buried
Figure 5.5 Carbon burial rate averaged from literature review (See Table 5.6) for each NLCD
category for: (A) Escambia County, FL; (B) Indian River County, FL; (C) Lafourche Parish, LA;
and (D) St. Landry Parish, LA. This is a metric for the valuation of a stable climate.
90
-------
Usable Water Value
Per hectare ($/ha/yr)
~ $16-$21
~ $22 - $25
^ $26 - $39
$40 - $54
^ $55 - $93
3 $94 - $189
| $190-$775
_| $776 $1,841
$1,842-$2,325
$2,326 - $3,098
Figure 5.6 Usable water value as calculated using average denitrification rates for each NLCD
category (See Table 5.7) for: (A) Escambia County, FL; (B) Indian River County, FL; (C)
Lafourche Parish, LA; and (D) St. Landry Parish, LA.
91
-------
($/ha/yr)
Figure 5.7 Usable air value as calculated using the average canopy cover percentage for each NLCD
category (See Table 5.8) for: (A) Escambia County, FL; (B) Indian River County, FL; (C)
Lafourche Parish, LA; and (D) St. Landry Parish, LA.
92
-------
Stable Climate Value
Per Hectare ($/ha/yr)
$0.00
I I $1 -$10
~ $11 -$20
^ $21 - $40
$41 - $52
| $53 - $58
| $59 - $65
| $66 -$117
$118-$253
$254 - $285
Figure 5.8 Stable climate value as calculated using average carbon burial rates for each NLCD
category (See Table 5.9) for: (A) Escambia County, FL; (B) Indian River County, FL; (C)
Lafourche Parish, LA; and (D) St. Landry Parish, LA.
93
-------
Flood Protection Value
Per Hectare ($/ha/yr)
~ o-1
I I 2 16
| 17-118
j 119-222
| 223 - 314
| 315 -432
| 433-493
| 494 - 781
| 782 - 942
I 943-1,578
Figure 5.9 Flood protection value as calculated using averaged curve numbers for each NLCD
category (See Table 5.10) for: (A) Escambia County, FL; (B) Indian River County, FL; (C)
Lafourche Parish, LA; and (D) St. Landry Parish, LA.
Differences in land cover among the four counties translates into differences in availability of
final ecosystem goods and services directly beneficial to humans. Mean usable water per hectare
was highest in Lafourche Parish, LA ($2,136/ha/yr), and lowest in Escambia County, FL
($865/ha/yr) (Figure 5.6; Table 5.7). Mean values for Indian River County and St. Landry Parish
were similar to one another and closer to Lafourche Parish than to Escambia County (Table 5.7).
Mean usable air per hectare was highest in Escambia County, FL ($369/ha/yr), and lowest in
Lafourche Parish, LA ($141/ha/yr) (Figure 5.7; Table 5.8). The increase in mean value for usable
air increased monotonically as a function of differences in canopy cover between the four
counties (Figure 5.3; Table 5.4). Lafourche Parish, LA, had the highest mean stable climate
value ($218/ha/yr), while Escambia County, FL, had the lowest ($95/ha/yr) (Figure 5.8; Table
5.9). For flood protection, Escambia County, FL, had the highest mean value ($412/ha/yr) and
94
-------
Lafourche Parish, LA, had the lowest ($52/ha/yr) (Figure 5.9; Table 5.10). The average per
hectare value of all of the ecosystem services - usable water, usable air, stable climate, and flood
protection - was highest for Lafourche Parish, LA ($637/ha/yr), and lowest for Escambia
County, FL ($435/ha/yr) (Table 5.11).
Table 5.7 Value ($/ha/yr) of maintaining water quality via natural denitrification by land cover
categories for four counties. Value is based on a literature review summarized in Appendix F: Table 5.
See text for details.
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Open Water
1,492.20
1,492.20
1,492.20
1,492.20
Open Space
127.80
127.80
120.60
120.60
Low Intensity
Development
91.80
93.60
93.60
97.20
Med Intensity
Development
54.00
55.80
54.00
52.20
High Intensity
Development
18.00
21.60
21.60
21.60
Barren
189.00
189.00
189.00
189.00
Deciduous Forest
39.60
39.60
39.60
39.60
Evergreen Forest
16.20
16.20
16.20
16.20
Mixed Forest
25.20
25.20
25.20
25.20
Shrub/Scrub
165.60
165.60
165.60
165.6
Grassland/Herbaceous
34.20
34.20
34.20
34.20
Pasture/Hay
775.80
775.80
775.80
775.80
Cultivated Crops
1,841.40
1,841.40
1,841.40
1,841.40
Woody Wetlands
3,097.80
3,097.80
3,097.80
3,097.80
Emergent Herbaceous
Wetlands
2,325.60
2,325.60
2,325.60
2,325.60
Total Average
864.82
1,737.50
2,135.90
1,883.51
Table 5.8 Value ($/ha/yr) of maintaining air quality via natural carbon processing in the canopy
cover. Values are given by land cover categories for four counties and are based on NLCD 2011 USFS
Tree Canopy cartographic data. See text for details.
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Open Water
0.08
0.09
0.00
0.04
Open Space
268.51
173.51
70.33
184.43
Low Intensity
Development
113.79
97.34
47.43
84.98
Med Intensity
Development
30.44
34.36
5.10
12.37
High Intensity
4.53
9.40
1.97
7.29
95
-------
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Development
Barren
40.68
33.92
9.94
92.68
Deciduous Forest
612.62
536.10
399.38
597.24
Evergreen Forest
587.17
391.96
248.56
578.35
Mixed Forest
587.82
240.69
456.02
584.12
Shrub/Scrub
431.46
187.48
160.54
413.95
Grassland/Herbaceous
175.97
80.88
77.71
133.90
Pasture/Hay
101.64
32.43
54.76
81.74
Cultivated Crops
35.12
65.36
28.88
48.48
Woody Wetlands
594.17
406.90
548.40
642.92
Emergent Herbaceous
Wetlands
215.99
98.42
25.79
330.45
Total Average
369.06
173.97
141.35
268.07
Table 5.9 Value ($/ha/yr) of maintaining stable climate via natural carbon burial. Values are given by
land cover categories for four counties and are based on a literature review summarized in Appendix F:
Table F.2. Citations: a is 2, 3, 4, 5, 6, 7; b is 1, 2.
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Open Water
139.82a
139.82a
284.38b
284.38b
Open Space
124.25
124.07
117.72
118.09
Low Intensity
Development
90.12
91.75
90.68
94.52
Med Intensity
Development
53.02
55.12
52.16
51.57
High Intensity
Development
17.16
20.26
20.23
20.37
Barren
0.00
0.00
0.00
0.00
Deciduous Forest
10.79
10.79
10.79
10.79
Evergreen Forest
63.84
63.84
63.84
63.84
Mixed Forest
37.32
37.32
37.32
37.32
Shrub/Scrub
0.00
0.00
0.00
0.00
Grassland/Herbaceous
40.77
40.77
40.77
40.77
Pasture/Hay
65.88
65.88
65.88
65.88
Cultivated Crops
58.88
58.88
58.88
58.88
Woody Wetlands
232.29
232.29
232.29
232.29
Emergent Herbaceous
Wetlands
253.42
253.42
253.42
253.42
Total Average
94.69
133.90
217.51
121.29
96
-------
Table 5.10 Value ($/ha/yr) of maintaining flood protection based on soil characteristics by land cover
categories for four counties. Value is based on Curve Number table in Zhang et al. 2011 and soil survey
data by SSURGO. The rate was based on a 30 year construction life cycle. See text for details.
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Open Water
0.00
0.00
0.00
0.00
Open Space
532.75
417.38
334.98
494.19
Low Intensity
Development
396.88
327.09
207.88
308.48
Med Intensity
Development
195.66
161.06
129.73
141.99
High Intensity
Development
118.36
104.98
87.24
94.12
Barren
15.74
15.74
15.74
15.74
Deciduous Forest
877.13
1,578.48
219.79
941.76
Evergreen Forest
825.32
604.71
306.99
780.52
Mixed Forest
963.05
586.32
224.19
918.85
Shrub/Scrub
421.55
313.72
377.80
432.27
Grassland/Herbaceous
493.15
441.71
309.12
511.51
Pasture/Hay
417.56
460.05
477.70
477.77
Cultivated Crops
207.21
221.67
209.49
278.91
Woody Wetlands
0.00
0.00
0.00
0.00
Emergent Herbaceous
Wetlands
0.00
0.00
0.00
0.00
Total Average
411.77
181.80
52.48
236.20
The differences in the four counties populations influence the per capita supply of each of the
ecosystem goods and services. The total values of ecosystem goods and services per capita were
calculated based on total area of each county by population (U.S. Census Bureau 2010). The total
usable water per capita value was highest in Lafourche Parish, LA ($6,679/person/yr) (Table
5.7), and lowest in Escambia County, FL ($503/person/yr). The total usable air per capita value
was highest in St. Landry Parish, LA ($782/person/yr) (Table 5.8), and lowest in Indian River
County, FL ($168/person/yr). For total stable climate per capita value, Lafourche Parish, LA, had
the highest ($680/person/yr) (Table 5.9) and Escambia County, FL, had the lowest
($55/person/yr). St. Landry Parish, LA, had the highest total flood protection per capita value
($689/person/yr) (Table 5.10) and Lafourche Parish, LA, had the lowest ($164/person/yr). The
average total per capita value of all of the ecosystem services - usable water, usable air, stable
climate, and flood protection - was highest for Lafourche Parish, LA ($7,965/person/yr) (Table
5.12), and lowest for Escambia County, FL ($l,012/person/yr).
97
-------
Table 5.11 Summary table of total value per hectare for select ecosystem goods and services for four
counties. Data are based on calculations by land cover category in previous tables. Mean total value is also
presented by community classification system group and by state. Group 1 includes Escambia and St.
Landry and Group 3 includes Indian River and Lafourche.
Escambia,
FL
Indian
River, FL
Lafourche,
LA
St. Landry,
LA
Type 1
Type 3
Florida
Louisiana 1
Usable Water
($/ha/yr)
864.82
1,737.50
2,135.90
1,883.51
1,374.16
1,936.70
1,301.16
2,009.71
Usable Air
($/ha/yr)
369.06
173.97
141.35
268.07
318.56
157.66
271.51
204.71
Stable
Climate
($/ha/yr)
94.69
133.90
217.51
121.29
107.99
175.70
114.30
169.40
Flood
Protection
($/ha/yr)
411.77
181.80
52.48
236.20
323.99
117.14
296.79
144.34
Total Average
($/ha/yr)
435.08
556.79
636.81
627.27
531.18
596.80
495.94
632.04
Table 5.12 Summary table of total value per capita per year for select ecosystem goods and services
for four counties. Data are based on calculations by land cover category in previous tables. Mean total
value is also presented by community classification system group and by state. Group 1 includes Escambia
and St. Landry and Group 3 includes Indian River and Lafourche. (U.S. Census, 2010).
Escambia,
FL
Indian
River, FL
Lafourche,
LA
St. Landry,
LA
Type 1
Type 3
Florida
Louisiana 1
Population
(capita)
297,619
138,028
96,318
83,384
381,003
234,346
435,647
179,702
Usable Water
($/person/yr)
503
1,682
6,679
5,494
1,595
3,736
877
6,129
Usable Air
($/person/yr)
215
168
442
782
339
281
200
600
Stable
Climate
($/person/yr)
55
130
680
354
120
356
79
529
Flood
Protection
($/person/yr)
239
176
164
689
338
171
219
408
Total
($/person/yr)
1,012
2,157
7,965
7,318
2,392
4,544
1,375
7,665
98
-------
The ecosystem goods and services largely vary by NLCD class, so the value distribution is similar to the
aforementioned spatial description for land cover. The usable water value is lower for development and
pastureland and higher in wetlands and open water. As a result, Escambia County has lower values
spread throughout the county; Indian River County has lower values along the eastern coastline, but
increased values in the middle and western portion of the county; Lafourche Parish has decreased values
along the narrow band of development, but higher values in the remaining area; St. Landry has less
usable water value in the developed pockets, but higher values in the surrounding areas (Figure 5.6;
Table 5.7).
Usable air is highest in areas of high canopy cover such as forest, and lowest in agriculture, wetlands,
and open vegetation. For Escambia County, the usable air values are low around its southeastern
development and northern agriculture, but are higher in its flanking woody wetlands and scattered
forests; Indian River's usable air values are low in its western open water and central agriculture, but
higher on its coastal woody wetlands; Lafourche Parish has high usable air values in its northern woody
wetlands and pastureland, but lower values in its surrounding open water and emergent herbaceous
wetlands; St. Landry has high values in its eastern and northern woody wetlands and lower values in its
agriculture and developed values found in the central and western portion (Figure 5.7; Table 5.8). Stable
climate is highest in wetlands and open water, and lowest in agriculture and open vegetation. For stable
climate value, in Escambia County, the southern and far eastern portions of the county have high values,
which lower values are scattered north; for Indian River, higher values are found vertically along the
eastern and western sections, with lower values in the agriculture in the middle; Lafourche Parish has
lower values along its development spine, surrounded by higher values in its wetlands; St. Landry has
high values in the eastern and northern parts of the Parish, with lower rates in the central and western
portion (Figure 5.8; Table 5.9).
Flood protection is focused in forested areas and lowest in wetlands. Flood protection value in Escambia
County is scattered throughout in its forests; Indian River has medium values in the agriculture and
development along the eastern half of the county, but has no protection value on the western side in the
wetlands and open water; Lafourche Parish has mid-level values along its central development string
with no flood protection in its surrounding water and wetlands; St. Landry has no protection value on the
western and northern sections and average values in the rest of it (Figure 5.9; Table 5.10).
The data were also organized by CCS (Chapter 2), geography, and state (Tables 5.11 and 5.12). For
CCS, Escambia County and St. Landry Parish are Type 1 communities, and Indian River County and
Lafourche Parish are Type 3 communities (See Chapter 2 for details). For geography, the two coastal
communities are Escambia and Indian River Counties (both in Florida) and the two inland communities
are Lafourche and St. Landry Parishes (both in Louisiana). The usable water value per hectare for Type
1 was less than for Type 3 and was less in the coastal communities than the inland ones. Type 1 had
higher usable air value per hectare than Type 3 as did the coastal communities to the inland ones. The
stable climate value per hectare was higher for Type 3 and inland communities than for Type 1 and
coastal communities. The flood protection value per hectare is over twice has high in both Type 1 and
coastal communities than in Type 3 and inland communities. Overall, the total average value of
ecosystem goods per hectare was higher in Type 3 and inland communities than in Type 1 and coastal
communities.
5.4 Discussion
Differences among communities in the production and availability of ecosystem services is a key factor
in defining community priorities to support decision making (Smith et al. 2013). Ecosystem goods and
services represent a community's ties to the local environment and as such contribute to economic
99
-------
stability, sense of place, and community identity (Smith et al. 2013). All EGS are potentially important,
but the four services considered here are expedient and relevant for coastal counties interested in
environmental sustainability and reducing the effects of coastal hazards and climate change (Barbier et
al. 2011). For that reason they are also useful for exploration of a central question, which is how well do
various delineations of communities inform about community priorities and therefore aid efforts to
inform the local decision process.
This study directly compared two delineations of community EGS value: U.S. state and coastal CCS
(See Chapter 2). The largest difference in EGS value between groups was for CCS with the exception of
useable water which differed more by U.S. state. Type 1 communities in both LA and FL had higher
specific value for usable air and flood protection, which is to be expected with an increase in canopy and
grass/shrub land associated with low and medium intensity development. Type 3 communities were
consistently lower in total area of both developed land and forest, and highest in wetlands, the latter
which provide higher denitrification but the former provide more carbon burial and water retention
during flood events (Pouyat et al. 2002, Ullah and Faulkner 2006).
These differences suggest tradeoffs exist between EGS categories in terms of benefits to humans. In the
abstract it seems plausible that flood protection, high denitrification, and high carbon burial could co-
exist at the spatial scale of this analysis (10-100 km), but in practice different land cover types
contributed to each and that land cover types were both distributed differently and affected differently
by human development linked to changes in impervious surface and canopy cover. Carbon burial, which
contributes to a more stable climate, and flood protection are clearly affected by development and
differences between CCS Groups 1 and 3 reflect this as these two groups differ principally in the level of
urbanization, which is higher in Type 1 communities. Denitrification, which contributes to clean water,
differed more by state than CCS group indicating a lower impact from development but a stronger
regional influence. These realized tradeoffs are important in that they can help clarify differences in the
impacts of development likely to affect decision outcomes. These trade-offs also support the conclusion
that local priorities for sustainability may differ based on the existing high value services they need to
sustain and/or improve and thus CCS groups can help inform the prioritization process. This is tied to
the notion that spatial demand for ecosystem services is the reciprocal of spatial supply (Burkhard et al.
2012).
The communities chosen for this study were selected to allow for a preliminary scoping comparison
between CCS and state level differences, which are useful for examining the utility of the CCS, but limit
drawing generalizations. This comparison was limited to four counties in two states, so conclusions
regarding CCS groups or states not considered here cannot be made. Nonetheless, differences in EGS
value by community type were observed and suggest a meaningful delineation can be made of EGS
services to communities. As with the human well-being index (Chapter 3) and stakeholder priorities
(Chapter 4), EGS resource availability and their inherent value to stake holders can be identified with
particular CCS groups and alongside these other characteristics, differences in EGS resources is an
important factor in decision support. The challenge moving forward will be to broaden the analysis to
more CCS groups and regions and examine the overall utility of this approach to classifying
communities.
100
-------
5.5 Literature cited
Barbier, E.B., S.D. Hacker, C. Kennedy, E.W. Koch, A.C. Stier, and B.R. Silliman. 2011. The value of
estuarine and coastal ecosystem services. Ecological Monographs 81.2:169-93.
Birch, M.B.L., B.M. Gramig, W.R. Moomaw, O.C. Doering III, and C.J. Reeling. 2011. Why metrics
matter: Evaluating policy choices for reactive nitrogen in the Chesapeake Bay watershed.
Environmental Science & Technology 45:168-74.
Boscolo, M., J.R. Vincent, and T. Panayotou. 1998. Discounting Costs and Benefits in Carbon
Sequestration Projects. Harvard Institute for International Development, Environment Discussion
Paper No. 41.
Burkhard, B., F. Kroll, S. Nedkov, and F. Mueller. 2012. Mapping ecosystem service supply, demand and
budgets. Ecological Indicators 21:17-29.
Homer, C., J. Dewitz, L. Yang, S. Jin, P. Danielson, G. Xian, J. Coulston, N. Herold, J. Wickham, and K.
Megown. 2015. Completion of the 2011 National Land Cover database for the conterminous
United States - Representing a decade of land cover change information. Photogrammetric
Engineering and Remote Sensing 81.5:345-54.
Millennium Ecosystem Assessment. 2005. Ecosystems and Human Well-Being, Island Press 5,
Washington, DC.
Murray, F.J., L. Marsh, and P. A. Bradford. 1994. New York State Energy Plan. Ed. Office, New York
State Energy. Vol 2. Albany, NY.
Pouyat, R., P. Groffman, I. Yesilonis, and L. Hernandez. 2002. Soil carbon pools and fluxes in urban
ecosystems. Environmental Pollution 116.1: S107-S118.
Russell, M., A. Teague, F. Alvarez, D. Dantin, M. Osland, J. Harvey, J. Nestlerode, J. Rogers, L. Jackson,
D. Pilant, F. Genthner, M. Lewis, A. Spivak, M. Harwell, and A. Neale. 2013. Neighborhood
Scale Quantification of Ecosystem Goods and Services. U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-13/295.
Smith, L.M., J.L. Case, L.C. Harwell, H.M. Smith, and J.K. Summers. 2013. Development of relative
importance values as contribution weights for evaluating human wellbeing: An ecosystem
services example. Human Ecology 41.4:631-41.
Ullah, S. and S.P. Faulkner. 2006. Denitrification potential of different land-use types in an agricultural
watershed, lower Mississippi valley. Ecological Engineering 28:131-40.
U.S. Census Bureau. 2010. Data derived from population estimates, American community survey, census
of population and housing, state and county housing unit estimates, county business patterns,
nonemployer statistics, economic census, survey of business owners, building permits. State and
County Quickfacts.
U.S. Department of Agriculture. 2015. Soil Survey Geographic Database (SSURGO). Natural Resource
Conservation Service.
101
-------
Wang, R., M.J. Eckelman, and J.B. Zimmerman. 2013. Consequential environmental and economic life
cycle assessment of green and gray stormwater infrastructures for combined sewer systems.
Environmental Science & Technology 47.19:11189-98.
Xian, G., J. Homer, J. Fry, N. Hossain, and J. Wickham. 2011. The change of impervious surface area
between 2001 and 2006 in the conterminous United States. Photogrammetric Engineering and
Remote Sensing 77.8:758-62.
Zhang, Y., Z. Zhang, S. Reed, and V. Koren. 2011. An enhanced and automated approach for deriving a
priori SAC-SMA parameters from the soil survey geographic database. Computers and
Geosciences 37:219-31.
102
-------
6 Synthesis
In this report communities have been compared based on four distinct metrics (community
classification, human well-being, stakeholder priorities, and availability of ecosystem goods and
services) with the purpose of seeking common ground for defining and measuring sustainability at the
local scale. Each metric can be interpreted independently as has been done in the respective chapters of
this report. However, the comparison of these metrics and more specifically the identification of
commonalities across metrics and between communities is the main goal of this report. This information
can be separated into two broad categories of useful information: how communities define sustainability,
and how that definition is tied to resources. Overlying this comparison is the final question of the
usefulness of the community type delineation for generalizing the findings to new communities.
Sustainability can be defined either subjectively by the community or objectively based on externally
derived metrics. In practice, the two can be tightly interrelated. The HWBI is a good example of an
objective measure of community quality, but it is not a measure of sustainability unless it measures what
community stakeholders wish to sustain. In the comparison of communities based on the HWBI, more
rural communities with a high degree of economic dependence on local natural resources had a lower
level of well-being (Chapter 3). This outcome was based on an unweighted objective measure (Smith et
al. 2013) and is highly consistent with similar outcomes from other studies (Cumming et al. 2014). Yet,
when stakeholder input was solicited on the subject of well-being, the domains were not found to be
consistently important in every community (Chapter 4). In fact, relative importance of the eight domains
of the HWBI, varied greatly in importance overall and varied among community types (Chapter 4). This
outcome suggests that well-being is not a constant feature that can be objectively measured the same
way in all communities, but must be weighted differently based on community characteristics. This
outcome is significant and limiting if the objective is a national comparison of well-being, but may still
be of value for community-level decision support. An alternative approach of understanding common
ground across communities will allow for a more targeted use of the HWBI at the local level.
Two major delineations of community type are considered here. First is geographic, or simply asking if
place (i.e., county, state) defines how communities measure well-being. The second was a community
CCS (Chapter 2) based on community demographics and economic dependencies. A comparison of
commonalities between communities was attempted in the final three chapters of this report involving
the HWBI, stakeholder weights on the HWBI, and available EGS resources. All three differed in
important ways but there was variance in the relative importance of geography versus community type.
The unweighted, objective HWBI varied more among community types than geographically. Similarly,
relative weightings of the eight domains of the HWBI differed by community type but some domains
were more strongly affected by geography so the outcome for weightings was domain dependent.
Finally, EGS resources were more strongly tied to geography with a smaller amount of variability
attributed to community type. The mixed result for the relative weightings and the dependence on
geography of EGS resources are related in that the domains of the HWBI vary in their association with
EGS resources and this may drive the relative importance of place in local weightings of the HWBI
domains (Chapter 5). Combined, these results suggest both geography and community type are
important for adapting a metric like the HWBI into a measure of local sustainability that is tied directly
to human benefits. They also suggest that a link between EGS resources and the HWBI is important and
should be quantified as a part of local decision support.
An additional critical question addressed in this report is the relative value of objective and subjective
information for measuring sustainability at the local scale. Many studies have compared human well-
103
-------
being or similar metrics across communities (Vemuri and Costanza 2006) or have developed purely
objective measures of sustainability such as ecological footprint (Mancini et al. 2016) or density of
green infrastructure (Van Mechelen et al. 2015). Yet, these objective studies come under criticism for
generalizing measures of benefit that are economically biased and therefore pre-determine well-being to
be something you have to purchase. This study examines the validity of this approach by asking
stakeholders in multiple communities what they value and prioritize (Chapter 4). The findings suggest a
moving target for measuring human benefit that is tied to tradeoffs in access to natural resources
(Chapter 5) and most importantly changes across the rural to urban gradient (Chapter 4). Therefore, a
balance between subjective and objective criteria in measuring sustainability at the local level may be
best achieved through use of the weighted HWBI.
A final important question for this report is how to make use of a community CCS in local decision
support. Community decision support is a national scale issue in that the collective impacts of multiple
local decisions can have large and pervasive results particularly in coastal areas. A good example is
watershed land use management where local decision making can impact water quality synergistically
and far down stream of the communities making the decisions. Central to the question of national- or
regional-scale community decision support is the balance between treating all communities the same or
focusing on the unique issues of each individual community. Treating all communities the same is not
recommended because it allows for avoidable variability in community characteristics to bias the
outcome and it may be viewed as 'externally driven', which limits the acceptability of the support by
stakeholders. In contrast, treating each community as totally unique is inefficient and ignores potentially
valuable commonalities. A key focus of this work has been to consider how this balance should be
struck in practice, and the outcome is that a community CCS can be a valuable way to approach the
issue. The community CCS examined in this report shows promise as a generalizing tool for decision
support and more importantly linking it to the HWBI allows for local input in a structured way, so that
the approach is transferable and adaptable as needed. Exploration of methods for effectively applying
the HWBI/CCS at the community level is therefore an important future research question.
The collective outcomes of this report strongly support exploration of a balanced approach for local
decision support that begins with identification of community type and the calculation of the weighted
HWBI. Questions remain about the optimal structure of the CCS and how well it can be applied in new
communities. This work will support new research and a coordinated case study that allows for
examination of this approach to measuring sustainability in multiple communities, and at the national
scale. Community-level decision support is a national scale issue and should be approached from that
point of view. Doing so will maximize the impact of EPA-led efforts and can result in a more effective
and accepted measure of community sustainability.
104
-------
6.1 Literature cited
Cumming, G.S., A. Buerkert, E.M. Hoffmann, E. Schlecht, S. von Cramon-Taubadel, and T. Tscharntke.
2014. Implications of agricultural transitions and urbanization for ecosystem services. Nature
515:50-57.
Mancini, M.S., A. Galli, V. Niccolucci, D. Lin, S. Bastianoni, M. Wackernagel, andN. Marchettini.
2016. Ecological footprint: Refining the carbon footprint calculation. Ecological Indicators
61:390-403.
Smith, L.M., J.L. Case, H.M. Smith, L.C. Harwell, and J.K. Summers. 2013. Relating ecosystem services
to domains of human well-being: Foundation for a US index. Ecological Indicators 28:79-90.
Van Mechelen, C., T. Dutoit, and M. Hermy. 2015. Adapting green roof irrigation practices for a
sustainable future: A review. Sustainable Cities and Society 19:74-90.
Vemuri, A.W. and R. Costanza. 2006. The role of human, social, built, and natural capital in explaining
life satisfaction at the country level: Toward a National Well-Being Index (NWI). Ecological
Economics 58:119-133.
105
-------
Appendix A
Vero Beach Community Workshop:
"Be the Spark - Help Shape the Growth and Character of Vero's Downtown"
Date
Tuesday, February 24, 2014
Location
United Way
[Insert Address]
9:00 AM- 3:00 PM (Check in at 8:30 AM)
Purpose
S Take a "fresh look" at our vision for the downtown by identifying what makes
Vero Beach a great place to live and visit and by identifying core community
values to help guide current initiatives and our long-term investments in the
downtown
J Contribute to a U.S. Environmental Protection Agency research project
identifying values that are important to different communities
8:30 AM
Sign In and Getting Settled
9:00 AM
Welcome and Introduction
• Welcoming Remarks (Community Leader and EPA Representative)
• Participant Introductions
• Agenda Review
9:30 AM
Building the Foundation: Vero Beach Community-Wide Values
Warm Up How Would You Describe Your Community?
Working Session What Do You Value in a Community?
Part 1 Breakout Group Discussion: When you think aboutyour
community, what do you care about most?
Part 2 Breakout Group Exercise: What values are important to this
community?
Part 3 Prioritization Exercise: Which of these community-wide goals do
you view as most important to Vero Beach?
12:00 PM
Exploring the Central Issue
Break and Working Lunch
Working Session "Be the Spark - Help Shape the Growth and Character of
Vero's Downtown"
Context: Presentation by local leader
Part 1 Breakout Group Exercise: When you think about an ideal vision
for downtown, what would that look like? What should we avoid?
Part 2 Breakout Group Exercise: What do we recommend?
What are the most important actions for the community to pursue
to meet our vision of Vero Beach's downtown in the near term?
Over the longer term?
2:30 PM
Final Thoughts (Next Steps, Wrap Up)
3:00 PM
Adjourn
106
-------
Appendix B
Table B.l List of goals.
Goal
Code
Category
Basic educational knowledge and skills
©
Education
Positive social and emotional development
(ED2)
More advanced knowledge and skills
(em)
Reasonable life expectancy
Health
Physical and mental well-being
©
Emotional well-being
Good quality healthcare
Healthy lifestyle and behavior
©
Enough time devoted to leisure activities
©
Leisure Time
Enough time devoted to physical activity and vacation
©
Reasonable time spent working and caring for others
©
Ability to afford basic necessities
©
Living Standards (Economics)
Reasonable income
©
Reasonable wealth
©
Job stability and satisfaction
©
Being safe
©
Safety and Security
Feeling (and being) safe
©
Resilience to hazards
©
Connectedness to nature
©
Connection to Nature
Cultural fulfillment
©
Cultural Fulfillment
Healthy family bonding
©
Social Cohesion
Supportive network of friends and family
©
Regular participation in community activities
©
Responsible engagement in our democracy
©
Satisfaction with others and the community
©
107
-------
Table B.2 Detailed explanation of goals.
Goal
What does it mean?
What can the community can do? (examples)
Education
/¦ \ Basic educational
knowledge and skills
We all need basic knowledge and skills to be able to
take care of our basic needs and give us a good start
on the path toward our future
Provide a good quality grade school education by
investing in our schools, our teachers, and our
children's participation in school
/ \ Positive social and
^ emotional
development
Being able to participate in society requires more than
just formal education; it means learning how to have
healthy social and emotional interactions with others
Provide social support services to help young parents
and foster safe and healthy home; provide safe and
healthy schools and after-school youth programs
/" \ More advanced
knowledge and skills
A basic education can get us started, but sometimes
we need more to fully develop our potential and
participate in society as parents, neighbors, workers,
community volunteers, etc.
Provide a good quality high school education that
meets the needs of all students; help high school
graduates start their career path; and support adult
education and job training
Health
/ \ Reasonable life
HL1
V_y expectancy
We all hope to live long, productive lives, and we hope
that our lives and the lives of those who we care
about won't be cut short by a terminal illness
Provide safe and affordable places to live; support
community clinics; provide public transit between
residences, hospitals, and other medical facilities
/" \ Physical and mental
well-being
To get the most out of life, we need good health and
freedom from debilitating physical and mental
illnesses like heart disease, cancer, diabetes, asthma,
obesity, and depression
Provide walkable, bike-friendly neighborhoods and
streets, parks for exercise and play, and affordable
recreation and exercise programs; provide public
transit between residences, clinics, and medical offices
/" \ Emotional well-being
(hlb)
Good health is more than just physical and mental
well-being; it means that we feel good about
ourselves and are satisfied with our lives
Provide safe, affordable places to live; provide safe
streets and parks; foster diverse public and private
cultural opportunities (e.g., fairs, music, restaurants)
/" "\ Good quality
fe) healthcare
Access to a family doctor and to good hospitals and
other healthcare facilities and providers helps us stay
healthy
Support community clinics; help recruit and retain
private family medical practices; and provide public
transit between residences, clinics, and medical offices
/ \ Healthy lifestyle and
behavior
Our health starts with us and our ability to make and
help others make healthy choices, like eating well and
avoiding behaviors like smoking and excessive drinking
Provide educational programs and social services for at
risk residents; provide safe and healthy schools and
after-school youth programs
Leisure Time
Enough time devoted
\ J to leisure activities
Leisure time—socializing with friends and family,
enjoying group sports and recreation, and just
relaxing—helps us refresh and get more out of life
Provide safe neighborhoods and parks for people to
exercise and play; provide affordable recreation and
exercise programs; ensure reasonable access to good
quality, affordable healthcare
/ \ Enough time devoted
\ / to physical activity
and vacation
Physical activity, such as running, walking, and
gardening, and getting away on vacation and to visit
family and friends helps us recover from stress and
maintain a positive emotional outlook
Provide safe neighborhoods and parks; affordable
recreation and exercise programs for all ages; provide
good access to nearby travel destinations, interstate
highways, and airports
Reasonable time
\ J spent working and
caring for others
Our days can be filled with work—work that we get
paid for and work as caregivers, e.g., for older family
members—but, if we over-commit, we become
drained, less productive, and less able to provide care
Support affordable transportation and housing (to
lessen demand on income); provide social services to
assist care-givers; provide affordable access to day care
and elderly care facilities
Living Standards (Economics)
/"\ Ability to afford basic
(EC1)
V_y necessities
Well-being starts with having enough resources to
meet our basic needs, including shelter, food, and
clothing
Support economic development and job opportunities;
support affordable transportation and housing; support
adult education and job training
Reasonable income
fe)
Life is more than survival; we need enough income
(from work, social security, etc.) to afford healthcare,
get to work, improve our prospects, give a gift, take a
vacation, etc.
Support economic development and job opportunities;
provide safe streets (for pedestrians, bikes, and cars)
and public transit between housing and job centers;
support post-secondary education and job training
Reasonable wealth
fe)
Our "wealth" is the sum of our assets; for many, this is
our home equity (value minus mortgage); wealth gives
us options, e.g., when someone gets sick, and allows
us to leave something for future generations
Provide social services to assist residents in addressing
credit issues and accessing down payment and
mortgage assistance programs
/" \ Job stability and
satisfaction
Job stability and satisfaction give us confidence in the
future and allow us to make smart decisions about
how we spend and save so we can afford the things
we need, now and in the future
Support diverse economic development and job
opportunities; work with local businesses and provide
education and job training to meet current and future
needs
108
-------
Safety and Security
/' \ Being safe
fe)
A safe environment is critical for living long, healthy,
productive lives; if we are not safe from crime,
accidents, and other hazards, it not only threatens to
shorten our lives, it can affect our ability to learn,
work, and enjoy life
Invest in strong police, fire, and emergency medical
services; enforce safe building codes; provide safe
streets for pedestrians, bikes, and cars; implement
community-wide emergency preparedness and
response plans
/ \ Feeling (and being)
® safe
If our surroundings don't feel safe—regardless of
whether they are or not—it has very real effects on
our lives and our ability to get out and go places, visit
neighbors, and enjoy our surroundings
Provide clear evidence of safety measures to ensure
safe neighborhoods and streets (e.g., police patrols,
street lights); provide community with crime statistics;
support community watch programs
Resilience to hazards
fe)
Security means being able to cope with hazards (e.g.,
natural disasters) when they occur; it means being
able to sustain yourself and your family and quickly
get back to normal to limit the impact of the event
Support community-wide natural disaster planning,
preparedness, and response; provide educational
resources for preparedness
Connection to Nature
Connectedness to
(NT1) „
V_y nature
We have an innate emotional connection with other
living organisms; experiencing this connection
improves our sense of emotional and spiritual well-
being
Provide safe and accessible public parks; offer
affordable nature programs; and support businesses
that provide opportunities for people to experience
nature
Cultural Fulfillment
/"N Cultural fulfillment
fe)
Our connections with the "culture" of our ancestors
and our existing community—including our shared
language, religion, cuisine, social habits, music, arts,
etc.—help us define "who we are"
Support community organizations, and community
events that celebrate different cultures; create mixed
use zones that support small shops and restaurants
with different cultural affiliations
Social Cohesion
/ Healthy family
bonding
The strength of our community fabric starts with a
strong family environment where we learn the
importance of spending time together and interacting
with others in a healthy, open-minded, and respectful
way
Provide social support services to help young parents;
support economic development and job opportunities
close to home; support affordable transportation and
housing
/^~Supportive network
\ J of friends and family
Our extended family and friends support us through
tough times; they also help us grow by allowing us to
confide our thoughts and feelings in others and by
helping us develop the character to be a trustworthy
friend to others
Provide walkable, bike-friendly neighborhoods and
parks for people to meet and enjoy; provide safe
streets and transit among neighborhoods; support
businesses that provide gathering places (e.g., music
venues, restaurants)
/"N Regular participation
\^/ in community
activities
When we participate in community activities, we meet
people with different experiences, racial and ethnic
backgrounds, and economic situations; we learn that
we all gain work together toward common goals
despite these differences
Provide public events (e.g., fairs, art exhibitions, music);
parks; affordable community programs arts, seniors,
and recreation programs to connect people with similar
interests; support businesses that provide gathering
places (e.g., music venues, restaurants)
/"N Responsible
engagement in our
democracy
Our government works for us, but only if we vote and
participating in public meetings; when we get
involved, we learn how different community needs
are considered, how decisions are made, and why our
voice matters
Conduct active outreach to encourage participation in
elections, referenda, and public meetings; provide
opportunities for citizens to participate on boards and
commissions; regularly communicate with citizens
regarding government activities and results
Satisfaction with
V^y others and the
community
When feel a sense of belonging, when we feel that
others care about our views and our needs, and when
we are motivated to help others feel the same, we
know that our efforts to build a strong community are
working
Support educational curriculum and cultural activities
that highlight the value of diversity and encourage
inclusiveness of all community members; encourage
participation of all segments of the community in
government decision-making
109
-------
Counties included in keyword search.
County
State
Brevard
FL
Broward
FL
Calcasieu
LA
Clay
FL
Dixie
FL
East Feliciana
LA
Escambia
FL
Fairhope
AL
Franklin
FL
Gadsden
FL
Gilchrist
FL
Glades
FL
Gulf
FL
Hancock
MS
Hardee
FL
Hendry
FL
Holmes
FL
Indian River
FL
Jackson
FL
Jackson
MS
Jefferson
FL
Lafourche
LA
Lake
FL
Levy
FL
Live Oak
Martin
FL
Mobile
AL
Monroe
FL
Moss Point
MS
Ocala
FL
Appendix C
County
State
Okaloosa
FL
Okeechobee
FL
Osceola
FL
Pascagoula
MS
Pearl River
MS
Pensacola
FL
Putnam
FL
Santa Rosa
FL
St. Johns
FL
Stone
MS
Sumter
FL
Suwannee
FL
Thomas
GA
Volusia
FL
Wakulla
FL
Walton
FL
Galveston
TX
Jefferson
LA
Manatee
FL
Nassau
FL
St. Charles
LA
St. James
LA
St. Landry
LA
Tangipahoa
LA
Taylor
FL
Vermilion
LA
West Baton
Rouge
LA
West Feliciana
LA
110
-------
Appendix D
The keyword list used to create the automated read of the community planning documents in Chapter
4.2 is located in the tables below. The list is broken down by domain and indicator, each with their own
set of "include", "near", and "exclude" words. The include column contains the primary set of words
used to describe each domain and indicator. Phrases were extracted if they contained one of the include
words. Near words were used to further specify the include words, generally verbs to identify phrases
that involved future actions and policy planning. With the inclusion of near words, phrases were
extracted only if they contained a word in the "include" and "near" columns. Phrases containing an
exclude word were ignored by the automated read.
A few notes on the formatting of the words in the tables below. The "|" symbol denotes "or", for
example, Natur|environment in the include column searched for phrases containing natur OR
environment. Some words were not the full spelling of the word, such as having "natur" instead of
"nature", this was to allow for variations of the word. "Natur" allowed for a more complete list,
including both natural and nature, whereas using "nature" excluded natural. "Hunt" pulled phrases that
included hunt, hunting, hunters, or hunts. If you just wanted "hunt" without any variations, a blank space
would be placed after the word, "hunt". Spaces placed before a word removed any prefix or other
variations. This was particularly useful for words like "art" and "create" to remove instances which they
are in larger words, "department" or "recreate".
Table D.l Keyword list for connection to nature.
Connection To Nature
Biophil ia
include
near
exclude
nature |
protect
resource | business | work | econom | hous | agriculture |
environment!
| support |enhanc|
landuse | home | recreat | resident | histor | pedestrian |
natural beauty
improv| preserve |
landscap | transportation | infrastructure |
preserving |
hazardous | waste | safe | urban | department
protected |
of | environmental impact study | protection | living
protecting | protects
environment
trail | hike |
support | enhance |
trailer | heritage | facilit | campaign | campus | outdoor
camp | canoe
build |improv|provi|
seating | encampment | life support | potable water
| kayak| fish | hunt|
establish | promot |
outdoor
creat|encourag|
expand | maintain |
attract
spiritual
environment | wild |
beaut|green|natur
sustainab
biodiversity
enhanc|improv|
increas|decreas| reduc|
creat | maintain
wildlife
expand | build |
enhanc | maintain |
creat |preserv|
protect
facilit | natural resource
zoo | aquarium
visit | recreat | attract |
provid | promot
Ill
-------
Table D.2 Keyword list for cultural fulfillment.
Cultural Fulfillment
Activity Participation
include
near
exclude
national park
support| enhance | build | improv| provid | establish | promot|
creat | encourag | expand | maintain | visit | develop
acre
heritage | cultur
expand | build | enhanc | maintain |
corridor | trail | art |
creat | preserve | protect | promot | support | attract | preserving
arts | work|indust|
promotion
entertainment |
support | promot | creat | provid | increas | enhanc | encourag |
state of the art | arts
fair | fairs
incorporat | integrat | recogniz | advocat | attract | attacts
degree | unfair | fair
| festival | art |
housing|fair
arts | concert |
share |fairhope |
fair | fairs
fairly
museum
preserv | operat | establish | promot | advertis | enhanc | explor |
ensur|visit
religio
assist | provid | promot | encourag
discrimination |
homeless
church | faith |
social | participat | attend | develop | provid | support | built |
churchill | church
spiritual
promot | offer | ensur
street | residential |
development
code|good
faith | development |
developed
cultural center
promot| preserv| support| establish | expand | creat| maintain
112
-------
Table D.3 Keyword list for education, basic knowledge.
Education
Basic Educational Knowledge and Skills of Youth
include
near
exclude
educat
provid | promot | enhanc |
history | department of education | higher | property |
improv|encourag|
audobon | home | government | financial | homeowner |
support | expand |
property | landlord | land use | training | drainage | air
maintain | creat | establish |
quality | energy | business | water | job | work | waste
upgrad
librar
maintain | promot | offer |
establish |coordinat|
encourag|expand|
explor |fund | support |
develop | upgrad | provide
educ| provides
educ | provide
servi | provides serv
public art | performing arts
math |
educat | teach | assist |
aftermath | mathew | neuroscience | marine
science |
learn | provid | promot |
science | spread
reading
encourag|expand|
enhanc
skill
teach | develop | provid |
assist | creat | support |
attract | promot | enhanc |
expand|educate |
educating | retain
training
assist | facilitat | promot |
job | employ | medical | hospital | staff | up to
support | offer |
date | business | financial | workforce
provid | teach|encourag |
enhance | improv
public education |
promote | program |
educated public |
provide
community education |
educated community
educate the public |
educate the community
113
-------
Table D.4 Keyword list for education, participation.
Education
Participation and Attainment
include
near
exclude
school
participat | attend | promot | support |
obtain | attain
school board | roadway | transportation |
economic | construct | siting | zoning |
parking | bike | bicycle | sidewalk | school
size | residential development | job | work
literacy
promot | provid | encourag | enhanc |
offer | support | train | prepar | improv |
fund | eradicat | reduc | decreas
lunch
degree|
graduat
increas | complet | trend | improv |
decreas
development | transportation | varying
college |
university
offer | course | program | educat |
provide | enhance | access | enroll |
opportunit
adult literacy |
adult education
promot | provid | encourag | enhanc |
offer | support | train | prepar | improv |
fund | eradicat | reduc | decreas
curriculum
educat | access | add | creat | design |
implement | establish |
encourag | support | promot | provid |
enhanc
Table D.5. Keyword list for education, development.
Education
Social, Emotional, and Developmental Aspects
include
near
exclude
social development | physical
development | emotional
development | social
support | emotional support | social
help | emotional
help | counsel | physical support
youth | young | child | adult | student | teach |
promot | educat | support
city | county |
counties | master
plan | forest |
agricultur|
homebuyer| home
owner
cognitive skill | mental development
youth | young | child | adult | student | teach | promot
| educat | support
fundamental
emotional well | emotional health
youth | young | child | adult | student | teach | promot
| educat | support
bully
youth | young | child | adult | student |
prevent | school | reduc | decreas | improv | counsel
114
-------
Table D.6 Keyword list for health, healthcare.
Health
Healthcare
include
near
exclude
healthcare |
provid | improv | access | enhanc | facilit | support | upgrad |
health care
promot
hospital
provid | build | increas | upgrad | improv | enhanc | support |
hospitality | government |
encourag | promot | help | offer
agriculture
doctor
employ | attract | enhanc | improv | attend | increas | provid |
deliver
bill
nurse
employ | attract | enhanc | improv | attend | increas | provid |
nursery | county | city |
deliver
parish | nurseries
medical facilit |
provid | build | increas | upgrad | improv | enhanc | support |
waste | shelter | fire |
medical service |
encourag | promot | help | offer
emergency management
medical assist |
agency
emergency |
clinic
Table D.7 Keyword list for health, personal well-being.
Health
Personal Well-being
include
near
exclude
well-being | well
improv | increas |
environment
being
rais | promot | support | enhanc | encourag | support | protect | provid
life satisfaction
improv | increas |
rais | promot | support | enhanc | encourag | support | protect | provid
happiness
improv | increas |
rais | promot | support | enhanc | encourag | support | protect | provid
115
-------
Table D.8 Keyword list for health, physical and mental health conditions.
Health
Physical and Mental Health Conditions
include
near
exclude
physical health
reduc | increas | decreas | prevent | provid | trend | higher |
lower | promot | encourag | enhance | support
economic | city
mental health
support | provid | promot | assist | access | offer | help | enhanc |
treat
environmental
diabetes
reduc | increas | decreas | health | prevent | provid | trend |
higher| lower|treat | support| promot| enhanc| encourag
cancer
reduc | increas | decreas | health | prevent | provid | trend |
higher| lower|treat | support| promot| enhanc| encourag
depression
reduc | increas | decreas | health | prevent | provid | trend |
national | economic |
higher| lower|treat | support| promot| enhanc| encourag
business | natur | storm |
weather
heart
reduc | increas | decreas | health | prevent | provid | trend |
higher| lower|treat | support| promot| enhanc| encourag
heart of | heartbreak
stroke
reduc | increas | decreas | health | prevent | provid | trend |
higher| lower|treat | support| promot| enhanc| encourag
asthma
reduc | increas | decreas | health | prevent | provid | trend |
higher| lower|treat | support| promot| enhanc| encourag
disease
reduc | increas | decreas | health | prevent | provid | trend |
higher| lower|treat | support| promot| enhanc| encourag
wastewater
ensure | expand | monitor | maintain | upgrad | build | construct
buildings
treatment
|fund | establish | install
Table D.9 Keyword list for health, life expectancy.
Health
Life Expectancy and Mortality
include
near
exclude
mortality
reduc | increas | decreas | health | prevent | provid | trend | higher |
lower | treat | support | promot | enhanc | encourag
suicide
reduc | increas | decreas | health | prevent | provid | trend | higher |
lower | treat | support | promot | enhanc | encourag
life
expectanc
reduc | increas | decreas | health | prevent | provid | trend | higher |
lower | treat | support | promot | enhanc | encourag
water
treatment
death | fatal
fewer | reduction | reduce | decrease | lower | growth | increas | prevent | support |
promot
manatee
116
-------
Table D.10 Keyword list for health, lifestyle.
Health
Lifestyle and Behavior
include
near
exclude
lifestyle
health | improv | enhanc | encourag | promot | support |
provid | retain | protect | preserv
behavior
health | safe | promot | enhanc | encourag | prevent
behavioral health
exercis | fitness
support | enhance | build | improv | provid | establish |
ranking | SWOT |S.W.O|
promot | creat | encourag | expand | maintain | increas |
rights | zoning | voting |
develop | developing
authority
smoking
prevent | encourag | educat | promot | reduc | increas |
decreas | enhanc | support
pregnancy
prevent | encourag | educat | promot | reduc | increas |
decreas | enhanc | support
alcohol
addict | prevent | health | behavior | risk | program | rehab |
service | support | enhanc | encourag | offer
public health
improv | enhanc | increas | promot | protect | ensure
117
-------
Table D.ll Keyword list for leisure time, activity participation.
Leisure Time
Activity Participation
include
near
exclude
leisure
support | enhanc | build | improv | provid |
establish | promot | creat | encourag | expand |
maintain | increas | activit
vacation
promot | encourag | enhanc | provid | access |
draw| attract | visit
public right
physical activit
encourag | exercis | increase | decreas |
reduc | bike | bicycl | walk | jog | run | health |
lifestyle | promot | provid | access | enhanc
construction
play | basketball |football |
establish | build | install | enhanc | improv |
display | played | role | government
soccer | tennis | volleyball |
invest | construct | upgrad | sponsor | promot |
| hotel | econom | playhouse
golf | baseball | sport|
encourag | provid | participat
-sport | physical
activit | softball
jog | run |water-
promot | exercis | establish | install | program |
campaign | agricultur |
sport | winter-sport | water
develop | activit | event
encampment | campus | runoff |
sport | winter sport
run-off
youth | child | kid | teen
activ| camp | league |
educat | librar | income | housing |
sport | service | program | recreat
child care | senior
care | crim | active adult | active-
adult | homebuyer | agricultur |
family services | child
support | wastewater | health
service | public
health | homeless | child abuse
recreation | park
build | provide | fund | funding| funded |
housing | car | highway | road |
creat | encourag | enhanc | maintain |
vehicle | parking | park and
support | promot | establish
ride | park-and
ride | business | office | agricultur |
mobile
home | sewer | wastewater |
aquacultur| industrial | trailer
park
118
-------
Table D.12 Keyword list for leisure time, retired seniors.
Leisure Time
Retired Seniors
include
near
exclude
retire | retiring |
elderly | senior
attract | encourag | support | promot | establish |
recreat | protect | activit | maintain
hous| apartment |
transportation
| high school | habitat |
development right
Table D.13 Keyword list for leisure time, time spent.
Leisure Time
Time Spent [amount of time]
include
near
exclude
leisure time
increas | decreas | reduc | provid | offer | promot | encourag | ensur |
rais | maintain | improv | enhanc | establish | creat
socialize | relax
promot| provid | creat | encourag | increas | improv| decreas | limit
Table D.14. Keyword list for leisure time, working age adults.
Leisure Time
Working Age Adults
include
near
exclude
work week
increas | decreas | reduc | provid | offer | promot | encourag |
ensur| rais| maintain | improv|enhanc|establish | creat
long hours
increas | decreas | reduc | provid | offer | promot | encourage
|ensur| rais | maintain | improv| enhanc | establish | creat
working day | work day
increas | decreas | reduc | provid | offer | promot | encourag |
ensur| rais| maintain | improv|enhanc|establish | creat
work balance | life
balace | work-life | work
life
119
-------
Table D.15 Keyword list for living standards, basic necessities.
Living Standards
Basic Necessities
include
near
exclude
afford | cheap |
health | work | food | water | energy |
housing | house | rent | home |
low-cost
electricity
apartment | unit |
stormwater | storm water
affordable hous | housing
encourag | provid | offer |
census data | department of
afford | diverse hous | housing
enhanc | increas | decreas |
housing
divers | cheap hous | cost
establish |fund | creat | prepar| preserv|
hous | affordable rent | diverse
develop |educat|promot| support
home | workforce hous | housing
work | affordable
apart | apartment afford | diverse
apart | apartment divers
drinking water | potable water
develop | provid | ensur | protect
food
access | expand | provid | basic |
expens | cheap | assist |
local | security | healthy
%
basic necessit
living standard |
rais | improv | decreas |
standard of living
maintain | preserv | enhanc
hous | home |
access
accessory | afford | sewer | water |
apartment
street | pedestrian | internet |
transportation | wildlife | habitat |
playhouse | accessories |
homelesss | transit | guard
house | sidewalk | manufactur |
automobil
homeless
encourag | offer | enhanc | establish | fund
| funding | creat | prepar | preserv |
develop |educat|promot| support |
prevent
shelter
encourag | offer | enhanc | establish | fund
| funding | creat | prepar | preserv |
develop |educat|promot| support |
prevent | provid
retirement community |
provid | build | encourag | enhanc |
retirement housing | assisted
offer | creat | fund | develop
living | nursing home | retirement
| promot | support | establish
home
park and ride | carpool | ride
promot | encourag | provid | offer |
share | ride share | ride-share
fund | establish | support | expand
| implement | incentiv | coordinat
public transportation | public
promot | encourag | provid | offer | fund |
transit
establish | support | expand
| implement | incentiv | coordinat |
improv| develop | maintain
120
-------
Table D.16 Keyword list for living standards, income.
Living Standards
Income
include
near
exclude
income
increas | decreas | rais | lower | trend | high | reduc
% | percent | housing | median
Table D.17 Keyword list for living standards, wealth.
Living Standards
Wealth
include
near
exclude
mortgage
program | service | counsel | assist
debt
management | assist | counsel | cut | raise | increas | decreas | red
uc
wealth
increas | rais|improv|
creat | enhanc | improv | promot | decreas | reduc
commonwealth
Table D.18 Keyword list for living standards, work.
Living Standards
Work
include
near
exclude
Employment
1 job
increas | enhanc | recruit |
creat | train | improv | retain | support | promot | encourag | attract
address | % | untrain |
network
econom
diversity | retain | expand | enhanc | employ | promot | improv
economic development |
socioeconomic
economic
development
diversity | retain | expand | enhanc | employ | promot | improv | job
creat
121
-------
Table D.19 Keyword list for safety and security, actual safety.
Safety and Security
Actual Safety
include
near
exclude
safe
provid | improv | ensur | promot | safe
access | support | encourag | enhanc
transportation |
safeguard | sidewalk |
traffic
transportation | traffic
safe
hazard
protect | reduc | mitigat | safe
toxic
safe | minimiz | spill | reduc
sidewalk | crosswalk
repair | construct | upgrad | improv | provid | enhanc |
establish | safe | build | develop
crime | murder | robbery |
rape | assault |
violence
prevent | protect | reduc | increas | lower | stop | decreas
| improv | stop | enhanc | encourag
financial
assistance | financial
security | economic security
assist
| support | coordinat | promot | provid | offer | increas |
decreas | improv | enhanc
emergenc| disaster |
hurricane | storm |
flood | snow | blizzard |
drought | fire | explosion
prepar | prevent | mitigate |coordinate
Table D.20 Keyword list for safety and security, perceived safety.
Safety and Security
Perceived Safety
include
near
exclude
perceived safety
community safety
feel safe
neighborhood watch | crime
prevent | crime stop
122
-------
Table D.21 Keyword list for social cohesion, attitude towards community.
Social Cohesion
Attitude Towards Others and the Community
include
near
exclude
small town | charm |
preserv|protect|
protection
lifestyle | character
enhance | maintain |
embrac | retain
satisfaction |
develop |enhanc|
groundwater | housing | development |
belonging | pride |
promot| improv|
investment | money | monetary | median |
value | beautification
increas | decreas |
customer | township |
encourag|preserv
census | stream | recreation | home | house |
manufactur | property | land
value | agricultur
quality of living | quality of life |
enhanc| improv| maintain |
community value | value of the
attract | promot | creat | preserv
community
cohesive community |
enhanc | improv | more |
community cohesion |
increas | decreas
family cohesion
historic
expand | build | enhanc |
ship | indust | soil | lake | stream | wetland |
maintain |
archaeological | hydrology | habitat |
creat|preserve| protect|
resource | storm
promot |
water | record | building | traffic
support | attract
| encourag | preserving
sense of place
creat | identity | establish |
provid | maintain |
preserv | improv | enhance |
develop | conserv | support |
unify | strengthen | encourag
Table D.22 Keyword list for social cohesion, democratic engagement.
Social Cohesion
Democratic Engagement
include
near
exclude
voting | vote | election
increas | turnout | participat | more | decreas |
reduc|encourag
politic
outreach | volunteer | involv| support | assist |
encourag | promot
democracy | democractic
practic | support | promot | improv | increas |
engagement
enhanc|encourag
public participation | public
promot | enhance | encourage | improv | increas |
employ | literacy |
planning | engagement |
decreas | involve | involved
business | park |
workshop | participation | public
school | recreat|
engagement | civic | public
art
hearing | community organiz
123
-------
Table D.23 Keyword list for social cohesion, family bonding.
Social Cohesion
Family Bonding
include
near
exclude
family
entertainment | recreation | camp | park | play |
multi-family | single family | single-
bond | event | museum |festival |friendly | social |
family | home | households |
communit
dwelling | hous
social
maintain | enhanc | creat | support | encourag |
cohesion
increas | improv | promot
community
increas | decreas | more | less | reduc | improv |
cohesion
enhanc | encourag | promot | support
Table D.24 Keyword list for social cohesion, social engagement.
Social Cohesion
Social Engagement
include
near
exclude
volunteer
assist | participat | coordinat | creat | encourag | provid
| promot | enhanc | support
respond
community
improv | provid | creat | enhanc | encourag | establish |
data | hunting
gathering | gathering
expand | allow | support | promot
place | gathering
point | public
gathering | gathering
space
group activit |family
promot | support | fund | provid | creat | coordinat |
undevelop
activit | community
encourag | enhanc | develop | improv | establish |
activit
expand | offer
social engagement
promot | support | fund | provid | creat | coordinat |
encourag | enhanc | develop | improv | establish |
expand
undevelop
extracurricular
promot | support | fund | provid | creat | coordinat |
encourag | enhanc | develop | improv | establish |
expand | offer
undevelop
welcome center | visitor
develop | provid | inform | support |fund | creat |
center
enhanc | expand | offer | improv | establish
124
-------
Table D.25 Keyword list for social cohesion, social support.
Social Cohesion
Social Support
include
near
exclude
social support
promot | support | fund | provid | creat |
coordinat | encourag | enhanc | develop |
improv | establish | expand | offer
undevelop
friend | neighbor
support | assist | help | promot | enhanc |
encourag
neighborhood | environmental |
pedestrian-
friendly | neighboring |
friendship house | pedestrian
friendly | regulation | user-
friendly | business | ecofriendly |
eco-friendly
125
-------
Appendix E
R code for keyword analysis annotated.
The following is formatted R script for conducting a keyword analysis of a pdf or word formatted
document. Both the document and the keyword list are input files for the analysis. The output for this
analysis is organized according to the eight domains of the Human well-being index (HWBI) (Smith et
al. 2013). See Table 3.1 or Appendix B for details.
####This can change to your own folder names, as long as all needed files in
a single folder
####directory <- "M:\\Gmeccs\\Workshops\\Inferring\\"
directory <- "L:\\Priv\\Sustainable Community Projects\\CoorCaseStudy
Task\\IanKraussFiles\\Keyword Planning\\RScript\\WorkingFolder\\"
###This is the table of keyword synonyms
###First column is HWBI category, 2nd is name of domainfor service), 3rd
column is indicators, all additional columns are synonyms
x <- read.table( file = paste(
directory,"TestSynonyms.csv",sep=""),sep=",",header=TRUE,as.is=TRUE)
###First convert the pdfs file to a text file, using Save As in Adobe, they
need to be in the same directory as this code & synonyms list####
filenames<-list.files(path = directory, pattern = "txt", all.files =
FALSE,full.names = FALSE, recursive = FALSE)
####This will overwrite the existing file, so rename if it matters to you
write.table(cbind("Planning Doc","HWBI category","Domain or
Service","Indicators","Synonym","Near","Exclude","Line Number","Matched
Words","Text from Planning
Doc"),file=paste(directory,"HWBI ALL Synonym Hits.csv",sep=""), sep =
row.names=FALSE,col.names=FALSE,append=FALSE)
write.table(cbind("Planning Doc","HWBI category","Domain or
Service","Indicators","Line Number","Text from Planning
Doc"),file=paste(directory,"HWBI Doc Hits.csv",sep=""), sep =
row.names=FALSE,col.names=FALSE,append=FALSE)
for(k in 1:NROW(filenames)){ ##cycle through all the planning docs
y by lines <- readLines(paste(directory,filenames[k],sep=""))
y whole<-paste(y by lines,collapse=' ')
y<-unlist(strspllt(y_whole,"[\f] | [.]\\s I [;] I [:] I [•] I [\t] I [?] I [!]"))
for(j in l:NROW(x)){ ####cycle over each HWBI indicator
ind hits<-integer(0)
if(x[j,4]=="include"){
for(i in 5:NCOL(x)){ ###cycle over each synonym (skipping
first four cols)
if(! (is.na(x[j,i])==TRUE | x [j,i]=="")) { ##if
synonym is not blank or NA then proceed
hits split<-NA
126
-------
near word="excl word="
hits<-grep(x[j,i], y,ignore.case=TRUE)
if(x[min(j +1,NROW(x)), 4 ]=="near") {if(x[j,3]==x[min(j +1,NROW(x) ),3] ) {if
(x[min(j +1,NROW(x)),i] !="") {
hits<-
intersect(hits,grep(x[j+1,i], y,ignore.case=TRUE))
near word=x[j+1,i];
}}}
if(x[min(j + 2,NROW(x)), 4 ]=="near") {if(x[j,3]==x[min(j +2,NROW(x) ),3] ) {if
(x[min(j +2,NROW(x)),i] !="") {
hits<-
intersect(hits,grep(x[j+2,i],y,ignore.case=TRUE))
near word=x[j+2,i];
}}}
if(x[min(j +1,NROW(x) ),4]=="exclude") {if(x[j,3]==x[min(j +1,NROW(x)),3])
{if(x[min(j +1,NROW(x)),i] !="") {
hits<-
setdiff(hits,intersect(hits,grep(x[j+1,i],y,ignore.case=TRUE)))
excl word=x[j+1,i];
}}}
if(x[min(j +2,NROW(x) ),4]=="exclude") {if(x[j,3]==x[min(j +2,NROW(x) ),3] )
{if(x[min(j + 2,NROW(x)),i] !="") {
hits<-
setdiff(hits,intersect(hits,grep(x[j + 2,i],y,ignore.case=TRUE) ))
excl word=x[j+2,i];
}}}
###find which words were hits
if(NROW(hits)>0){
hits split<-rep(0,NROW(hits))
y split<-strsplit(y[hits]," ")
for(w in 1:NROW(hits)){
if(!(near word=="")){hits split[w]<-
paste(y split[[w]][grep(gsub(" ","|",gsub(" gsub(","",gsub("
\\ I ", " Igsub ("\\ I
","|",paste(x[j,i],"|",near word,sep="")))))),y split[[w]],ignore.case=TRUE)
],sep="",collapse=", ")}
if(near word==""){hits split[w]<-
paste(y split[[w]][grep(gsub(" ","|",gsub(" gsub(","",gsub("
\\ I " Igsub ("\\ I
","|",x [ j,i]) )) )),y split[[w]],ignore.case=TRUE)],sep="",collapse=", ") }
}
}
127
-------
### this writes ALL matches for ALL synonynms,
comment out using ### if not needed
write.table(cbind(filenames[k],x[j,l],x[j,2],x[j,3],x[j,i],near word,e
xcl word,hits,hits split,y[hits]),file=paste(directory,"HWBI ALL Synonym Hit
s.csv",sep=""), sep =
row.names=FALSE,col.names=FALSE,append=TRUE,qmethod="double")
###remove duplicates for the indicator i
if(i==5){ind hits<-hits}
if(i>5){ind hits<-union(ind hits,hits)}
}
}
###this writes the lines associated with an indicator
(ignores which synonym they came from)
write.table(cbind(filenames[k],x[j,1],x[j,2],x[j,3],ind hits,y[ind hit
s]),file=paste(directory,"HWBI Doc Hits.csv",sep=""), sep =
row.names=FALSE,col.names=FALSE,append=TRUE,qmethod="double")
}
}
}
####Read back in the file you just created
WordMatches <- read.table( file = paste(
directory,"HWBI Doc Hits.csv",sep=""),sep=",",header=TRUE,as.is=TRUE,fill=TR
UE)
names(WordMatches)
######Change to a 1 or a 0 if there was a text match to a keyword
Hits<-rep(0,NROW(WordMatches))
Hits [ ! (WordMatches$Text.from.Planning.Doc=="")]=1
###Count the number of hits for Domains/Services
index<-
aggregate(x=l:NROW(Hits),by=list(WordMatches$Planning.Doc,WordMatches$HWBI.c
ategory,WordMatches$Domain.or.Service),FUN="min")
Hit Counts<-
aggregate(Hits,by=list(WordMatches$Planning.Doc,WordMatches$HWBI.category,Wo
rdMatches$Domain.or.Service),FUN="sum")
Hit YesNo<-Hit Counts$x
Hit YesNo[Hit Counts$x>0]=1
output<-cbind(Hit Counts,Hit YesNo)
output<-output[order(index$x),] ### this is to get back the original order
write.table(cbind("Planning Doc","HWBI Category","Domain/Service","Number of
Hits","Hit YesNo"),file=paste(directory,"HWBI Domain Service Hits.csv",sep=
""), sep = row.names=FALSE,col.names=FALSE,append=FALSE)
write.table(output,file=paste(directory,"HWBI Domain Service Hits.csv",sep="
"), sep = row.names=FALSE,col.names=FALSE,append=TRUE)
###Count the number of hits for Indicators
128
-------
index<-
aggregate(x=l:NROW(Hits),by=list(WordMatches$Planning.Doc,WordMatches$HWBI.c
ategory,WordMatches$Domain.or.Service,WordMatches$Indicator),FUN="min")
Hit Counts<-
aggregate(Hits, by=list(WordMatches$Planning.Doc,WordMatches$HWBI.category,Wo
rdMatches$Domain.or.Service,WordMatches$Indicator),FUN="sum")
Hit YesNo<-Hit Counts$x
Hit YesNo[Hit Counts$x>0]=1
output<-cbind(Hit Counts,Hit YesNo)
output<-output[order(index$x),] ### this is to get back the original order
write.table(cbind("Planning Doc","HWBI Category","Domain/Service","Indicator
","Number of Hits","Hit YesNo"),file=paste(directory,"HWBI Indicator Hits.cs
v",sep=""), sep = row.names=FALSE,col.names=FALSE,append=FALSE)
write.table(output,file=paste(directory,"HWBI Indicator Hits.csv",sep=""),
sep = row.names=FALSE,col.names=FALSE,append=TRUE)
###############
###Count the number of hits for Domains/Services, but remove duplicates
(each line counted only once for each domain)
noDups<-
aggregate(x=l:NROW(Hits),by=list(WordMatches$Planning.Doc,WordMatches$HWBI.c
ategory,WordMatches$Domain.or.Service,WordMatches$Line.Number),FUN="min")
Hits<-rep(0,NROW(WordMatches))
Hits[noDups$x]=1
index<-
aggregate(x=l:NROW(Hits),by=list(WordMatches$Planning.Doc,WordMatches$HWBI.c
ategory,WordMatches$Domain.or.Service),FUN="min")
Hit Counts<-
aggregate(Hits,by=list(WordMatches$Planning.Doc,WordMatches$HWBI.category,Wo
rdMatches$Domain.or.Service),FUN="sum")
Hit YesNo<-Hit Counts$x
Hit YesNo[Hit Counts$x>0]=1
output<-cbind(Hit Counts,Hit YesNo)
output<-output[order(index$x),] ### this is to get back the original order
write.table(cbind("Planning Doc","HWBI Category","Domain/Service","Number of
Hits","Hit YesNo"),file=paste(directory,"HWBI Domain Service Hits NoDups.cs
v",sep=""), sep = row.names=FALSE,col.names=FALSE,append=FALSE)
write.table(output,file=paste(directory,"HWBI Domain Service Hits NoDups.csv
",sep=""), sep = row.names=FALSE,col.names=FALSE,append=TRUE)
###Count the number of hits for Indicators, but remove duplicates (each line
counted only once for each domain)
noDups<-
aggregate(x=l:NROW(Hits),by=list(WordMatches$Planning.Doc,WordMatches$HWBI.c
ategory,WordMatches$Domain.or.Service,WordMatches$Indicator,WordMatches$Line
.Number),FUN="min")
Hits<-rep(0,NROW(WordMatches))
Hits[noDups$x]=1
index<-
aggregate(x=l:NROW(Hits),by=list(WordMatches$Planning.Doc,WordMatches$HWBI.c
ategory,WordMatches$Domain.or.Service,WordMatches$Indicator),FUN="min")
129
-------
Hit Counts<-
aggregate(Hits, by=list(WordMatches$Planning.Doc,WordMatches$HWBI.category,Wo
rdMatches$Domain.or.Service,WordMatches$Indicator),FUN="sum")
Hit YesNo<-Hit Counts$x
Hit YesNo[Hit Counts$x>0]=1
output<-cbind(Hit Counts,Hit YesNo)
output<-output[order(index$x),] ### this is to get back the original order
write.table(cbind("Planning Doc","HWBI Category","Domain/Service","Indicator
","Number of Hits","Hit YesNo"),file=paste(directory,"HWBI Indicator Hits No
Dups.csv",sep=""), sep = row.names=FALSE,col.names=FALSE,append=FALSE)
write.table(output,file=paste(directory,"HWBI Indicator Hits NoDups.csv",sep
=""), sep = row.names=FALSE,col.names=FALSE,append=TRUE)
130
-------
Appendix F
Table F.l References for denitrification rates.
NLCD Class
Citations
Open Water
An and Gardner 20021, An et al. 20012, Bianchi et al. 19993, Brenner et al. 20014, DeLaune
etal. 2005s, Fennell etal. 20096, Gardner etal. 20067, Gihring etal. 20108, Heffernan etal.
20109, James et al. 201110, Joye and Anderson 200811, Messer and Brezonik 198312,
Mortazavi et al. 200013, Pina-Ochoa and Alvarez-Cobelas 200714, Seitzinger 198815,
Seitzinger etal. 200616, Smith etal. 198517
Open Space
Barton etal. 199918, Raciti etal. 201119, Robertson etal. 198 720, Tsai 198921
Low Intensity Development
Barton etal. 199918, Raciti etal. 201119, Robertson etal. 198 720, Tsai 198921
Med Intensity Development
Barton etal. 199918, Raciti etal. 201119, Robertson etal. 198 720, Tsai 198921
High Intensity Development
Barton etal. 199918, Raciti etal. 201119, Robertson etal. 198 720, Tsai 198921
Barren
Walker etal. 199222
Deciduous Forest
Barton et al. 199918, Chestnut et al. 199 923 Henrich and Haselwandter 199724, Robertson
etal. 198725
Evergreen Forest
Barton et al. 199918, Henrich and Haselwandter 199 724, Robertson et al. 198725
Mixed Forest
Barton etal. 199918, Dutch and Ineson 199026, Goodread and Keeney 198427
Shrub/Scrub
Walker etal. 199222
Grassland/Herbaceous
Robertson et al. 198 725, Tsai 198928
Pasture/Hay
Barton et al. 199918, Espinoza 199 729, Hofstra and Bouwman 20 0 530, Seitzinger et al.
2006i6, Tsai 198928
Cultivated Crops
Pina-Ochoa and Alvarez-Cobelas 200714
Woody Wetlands
Bowden 198731, DeLaune et al. 199832, Gale et al. 199 3 33, Lindau et al. 200834, Seitzinger
199435, Walbridge and Lockaby 199436
Emergent Herbaceous Wetlands
Craft et al. 200937, DeLaune et al. 198938, Dodla et al. 200839, Gale et al. 199 3 33, Morris
199 140, Nixon and Lee 198641, Reddy et al. 198942, Seitzinger 198 743, Wigand et al. 200444
Table F.2 References for carbon burial rates.
NLCD Class
Citations
Open Water
Brennar etal. 20011, Craft and Richardson 19932, Downing etal. 20083, Duarte et al.
20054, Duarte et al. 20075, Gacia et al. 20026, McCleod et al. 20117
Open Space
Pouyat etal. 20098, Pouyat etal. 20109, Pouyat et al. 201110, Qian and Follett 200211,
Raciti et al. 201112
Low Intensity Development
Pouyat etal. 20098, Pouyat etal. 20109, Pouyat etal. 201110, Qian and Follett 200211,
Raciti et al. 201112
Med Intensity Development
Pouyat etal. 20098, Pouyat etal. 20109, Pouyat etal. 201110, Qian and Follett 200211,
Raciti et al. 201112
High Intensity Development
Pouyat etal. 20098, Pouyat etal. 20109, Pouyat etal. 201110, Qian and Follett 200211,
Raciti et al. 201112
Barren
N/A
Deciduous Forest
Downing et al. 200813, Johnston et al. 199614, Laffoley and Grimsditch 200915, McCleod et
al. 20117
Evergreen Forest
Garten Jr. 200216, Hooker and Comptoon 200317, Hooker and Comptoon 200418,
Huntington 199519, Richter et al. 199 9 20, Schiffman and Johnson 198921
Mixed Forest
Downing et al. 200813, Garten Jr. 200216, Hooker and Comptoon 200317, Hooker and
Comptoon 200418, Huntington 199519, Johnston etal. 199614, Laffoley and Grimsditch
200915, McCleod etal. 20117, Richter etal. 199 920, Schiffman and Johnson 198921
Shrub/Scrub
N/A
Grassland/Herbaceous
Burke et al. 199 522, Downing et al. 200813, Gebhart et al. 199423, Knops and Tilman
200024, Laffoley and Grimsditch 200915, Post and Kwon 199925
Pasture/Hay
Downing et al. 20013
Cultivated Crops
Heath and Pacala 200126, Houghton et al. 199927
Woody Wetlands
Briethaupt et al. 201228
Emergent Herbaceous Wetlands
Brennar et al. 200129, Chmura et al. 200230, Day et al. 200431, DeLaune et al. 198 1 32,
Downing et al. 200813, Duarte et al. 200533, Laffoley and Grimsditch 200915
131
-------
Appendix G
Table G.l NLCD category descriptions (Homer 2015).
NLCD Code
NLCD Description
11
Open Water
21
Open Space
22
Low Intensity Development
23
Med Intensity Development
24
High Intensity Development
31
Barren
41
Deciduous Forest
42
Evergreen Forest
43
Mixed Forest
52
Shrub/Scrub
71
Grassland/Herbaceous
81
Pasture/Hay
82
Cultivated Crops
90
Woody Wetlands
95
Emergent Herbaceous Wetlands
132
-------
Appendix H
Table H.l Percentage of each soil type by NLCD land cover categories for the four counties. This is used to calculate a weighted Curve Number value
that is a metric for Flood Protection.
NLCD
Escambia, FL (%)
Indian River, FL (%)
Lafourche, LA (%)
St. Landry, LA (%)
A
B
C
D
A
B
C
D
A
B
c
D
A
B
C
D
11
88.23
7.36
4.41
0.00
95.92
0.00
3.91
0.18
97.07
0.00
0.28
2.65
93.36
0.21
0.91
5.52
21
71.80
15.97
12.23
0.00
64.46
2.09
26.58
6.87
54.50
0.00
0.47
45.03
71.61
7.34
13.97
7.08
22
86.86
5.29
7.86
0.00
71.82
3.71
16.53
7.94
44.00
0.00
0.74
55.26
65.78
5.53
20.32
8.38
23
86.62
4.05
9.33
0.00
66.58
3.66
22.22
7.54
52.83
0.00
1.30
45.88
52.49
7.51
27.50
12.50
24
87.62
1.83
10.55
0.00
71.34
3.60
18.98
6.07
56.38
0.00
0.07
43.55
58.12
3.26
26.74
11.89
31
92.57
3.81
3.61
0.00
60.81
0.27
36.31
2.61
88.88
0.00
2.40
8.72
81.84
0.94
13.21
4.01
41
75.08
14.65
10.27
0.00
69.74
0.00
25.88
4.39
8.01
0.00
0.00
91.99
80.47
11.28
7.79
0.46
42
69.06
20.90
10.04
0.00
65.32
0.42
4.85
29.41
26.29
0.00
7.73
65.98
70.05
11.33
17.34
1.28
43
80.40
14.08
5.51
0.00
64.30
0.59
1.17
33.94
9.17
0.00
0.00
90.83
78.26
13.39
7.93
0.43
52
65.49
23.80
10.71
0.00
58.52
0.75
10.18
30.55
72.99
0.00
1.09
25.92
76.11
6.72
10.59
6.58
71
62.93
23.98
13.09
0.00
66.17
0.87
24.78
8.18
49.39
0.00
0.97
49.64
76.49
3.85
8.84
10.83
81
46.12
34.72
19.16
0.00
70.31
1.75
14.00
13.95
75.54
0.00
3.60
20.86
70.03
6.17
14.98
8.81
82
13.91
44.59
41.50
0.00
40.82
0.05
59.08
0.05
44.70
0.00
0.60
54.71
70.30
1.45
12.48
15.77
90
69.07
25.41
5.53
0.00
62.33
4.76
27.73
5.18
30.01
0.00
4.99
65.00
66.69
0.69
3.04
29.58
95
78.75
17.33
3.92
0.00
75.50
1.89
21.74
0.87
65.65
0.00
8.66
25.69
61.89
0.36
2.50
35.26
133
-------
Appendix I
Table 1.1 Curve numbers by land cover category and soil type. Curve Number (CN) is an index that represents
a land cover category's ability to hold water after a rain event has begun and just before runoff starts to occur. The
lower a CN value, the lower the runoff potential. For example, a land use type with a CN of 30 has a very high
water retention rate and a low runoff potential, whereas a land use type with a CN of 100 has a very low water
retention rate and a high runoff potential. A, B, C, and D are different soil types and textures distinguished by
varying combinations of sand, loam, silt, and clay. (Zhang et al. 2011).
NLCD
A
B
c
D
Open Water
100
100
100
100
Open Space
29
48
61
69
Low Intensity Development
40
56
67
74
Med Intensity Development
58
70
79
83
High Intensity Development
70
79
84
87
Barren
95
95
95
95
Deciduous Forest
19
39
53
61
Evergreen Forest
19
39
53
61
Mixed Forest
19
39
53
61
Shrub/Scrub
34
52
64
72
Grassland/Herbaceous
29
48
61
69
Pasture/Hay
29
48
61
69
Cultivated Crops
45
57
66
70
Woody Wetlands
100
100
100
100
Emergent Herbaceous Wetlands
100
100
100
100
134
-------
Table 1.2 Weighted curve number by land cover category for four counties assigned based on the soil
hydrogroup of each NLCD class. Curve Number (CN) is an index that represents a land cover category's ability
to hold water after a rain event has begun and just before runoff starts to occur. The higher the curve number, the
less able the soil is to absorb runoff. For example, in Escambia, FL, Deciduous Forest can hold almost four times as
much water as woody wetlands. (Zhang et al. 2011).
NLCD
Escambia, FL
Indian River, FL
Lafourche, LA
St. Landry, LA
Open Water
100.00
100.00
100.00
100.00
Open Space
35.95
41.74
47.16
37.70
Low Intensity Development
42.97
47.76
58.99
49.22
Med Intensity Development
60.44
64.99
69.74
67.80
High Intensity Development
71.64
74.01
77.41
76.06
Barren
95.00
95.00
95.00
95.00
Deciduous Forest
25.42
15.93
57.63
24.10
Evergreen Forest
26.59
33.09
49.34
27.70
Mixed Forest
23.69
33.77
57.15
24.55
Shrub/Scrub
41.50
48.80
44.18
40.89
Grassland/Herbaceous
37.75
40.37
49.17
36.89
Pasture/Hay
41.73
39.39
38.50
38.49
Cultivated Crops
59.07
57.43
58.80
51.74
Woody Wetlands
100.00
100.00
100.00
100.00
Emergent Herbaceous
Wetlands
100.00
100.00
100.00
100.00
135
-------
SEPA
United States
Environm,ental Protection
Agency
Office of Research and
Development (8101R)
Washington, DC 20460
EPA/600/R-16/178
------- |