Assessing Social Vulnerability to Flooding in Metro Manila Using Principal Component Analysis

Abstract

Earlier works measuring social vulnerability to flooding in cities focused mainly on biophysical, socio-demographic, or structural components of the hazard. Most of them are either: general, focusing on the city as a whole or specific, focusing on a particular district or community. These studies highlight the need to expand the concept of social vulnerability by examining the socioeconomic and political dimensions of urban vulnerability at both city and community levels through quantitative methodologies. This paper argues that the socioeconomic and political components of urban vulnerability tend to interact with the biophysical dimensions of flooding and tend to magnify/intensify its effects on the city and its residents, especially the poor residing along river lines. Using principal component analysis (PCA), this study measured the social vulnerability index of 17 Metro Manila cities/municipality based on the 2010 census data and a sample of flood prone communities drawn from the three flood basins of the metropolis. Results show that the components of social vulnerability vary across the city, barangay, and household levels, highlighting the multidimensionality and heterogeneity of social vulnerability across different local contexts. The expanded analysis also revealed that social capital and access to basic services are significant components of social vulnerability and adaptive capacities at the household level.

Full text

5 31IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMtQQ
+645*/$)"3-&4(4&&
"/%&.."&103*0
Assessing Social
Vulnerability to
Flooding in Metro
Manila Using Principal
Component Analysis
Earlier works measuring social vulnerability to flooding in cities focused mainly
on biophysical, socio-demographic, or structural components of the hazard.
Most of them are either: general, focusing on the city as a whole or specific,
focusing on a particular district or community. These studies highlight the need
to expand the concept of social vulnerability by examining the socio-economic
and political dimensions of urban vulnerability at both city and community
levels through quantitative methodologies. This paper argues that the socio-
economic and political components of urban vulnerability tend to interact with
the biophysical dimensions of flooding and tend to magnify/intensify its effects
on the city and its residents, especially the poor residing along river lines. Using
principal component analysis (PCA), this study measured the social vulnerability
index of 17 Metro Manila cities/municipality based on the 2010 census data and
a sample of flood prone communities drawn from the three flood basins of the
metropolis. Results show that the components of social vulnerability vary across
the city, barangay, and household levels, highlighting the multidimensionality
and heterogeneity of social vulnerability across different local contexts. The
expanded analysis also revealed that social capital and access to basic services
are significant components of social vulnerability and adaptive capacities at the
household level.
Keywords: social vulnerability, principal component analysis, climate
change adaptation, riverine communities, urban development

5 4 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Her World Awashed. (Photo by John Javellana)

5 51IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PM
INTRODUCTION
The World Risk Report of 2014 ranks the Philippines as second
in the world in terms of risk (UNU-EHS 2014). This means
that there is a higher probability that its population will suffer
damages and losses from different environmental hazards such as floods,
typhoons, and earthquakes. Over the years, the number and scale of
natural disasters have been on the rise, and so are the potential increasing
impacts on human populations (Oliver-Smith 2008).
Although the concept of vulnerability appears in various disciplines
of applied research, there is really no common working definition for the
term (Birkmann and Wisner 2006). In fact, there are various definitions
of vulnerability in the literature, depending on conceptual models and
frameworks (Adger 1999; Downing, Olsthoorn, and Tol 1999; Cutter
1996). From the standpoint of this study, vulnerability is the susceptibility
of people to the harm caused by exposure to a hazard (Birkmann 2006;
Cutter, Boruff, and Shirley 2003). Thus, vulnerability affects the ability of
a population to respond to and recover from the hazard. The term social
vulnerability explicitly focuses on the demographic and socio-economic
factors that increase or attenuate the impacts of hazard events on local
populations (Cutter, Emrich, Webb, and Morath 2009; Tierney, Lindell,
and Perry 2001).
Cutter, Barnes, Berry, Burton, Evans, Tate, and Webb (2008) argues
that there is a shift from a qualitative work centered on conceptual
frameworks to quantitative and empirical measures of social vulnerability.
In order to understand social vulnerability in the United States, Cutter,
Boruff, and Shirley (2003) developed the Social Vulnerability Index
(SOVI) to quantify the demographic and socio-economic quality of
Justin Charles G. See is lecturer at the Department of Sociology and
Anthropology, Ateneo de Manila University and sociologist-statistician for
the social sector of the Coastal Cities at Risk (CCaR) project of the Manila
Observatory and the Ateneo de Manila University. Dr. Emma E. Porio is
Professor of Sociology at the Department of Sociology and Anthropology,
Ateneo de Manila University. She is a science research fellow at the Manila
Observatory, and principal co-investigator of the CCaR project. The authors can
be emailed at: justin_see2006@yahoo.com.

5 6 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
place. Using factor analysis, 42 variables derived from the US Census
were reduced to 11 statistically independent factors. These factors were
aggregated using a simple additive model in order to compute a summary
score (Cutter, Boruff, and Shirley 2003). The factors that were found
to be significant contributors to social vulnerability were different for
each county, highlighting the interactive nature of social vulnerability,
i.e. some components increase social vulnerability whereas some reduce
or moderate it (Cutter, Boruff, and Shirley 2003).
Adger (2006) views vulnerability as dynamic phenomenon often
in a constant state of flux. He further elaborates that “measurement of
vulnerability should reflect social processes and material outcomes… and
is therefore difficult to reduce to a single metric.” Nevertheless, Adger
argues that quantitative measures complement narratives of stakeholder-
led or qualitative assessments of vulnerability in places and contexts
(Adger 2006). It can also be used to validate and triangulate data to
derive more robust measures for both policy and intervention (Downing,
Butterfield, Cohen, Hug, Moss, Rahman, Youba, and Stephen 2001).
Birkmann (2006) argues that vulnerability is linked to society, economy,
and the environment. The “BBC Framework” explicitly illustrates how
the social, economic and environmental dimensions of human security
can be integrated with existing hazard and risk concepts. Furthermore,
Birkmann shows that within the context of disaster management, hazard,
and vulnerability are interconnected and altogether define risk. Thus, the
main structuring points for the development of vulnerability indicators
include components of exposure, susceptibility, and capacity (Birkmann
2006).
Using the insights of Adger (2006), and Birkmann (2006), Cutter,
Boruff, and Shirley (2003), this article explores the development of
a quantitative, empirical, and multi-dimensional measure for social
vulnerability in Metro Manila. It seeks to identify the statistically
significant socio-economic and demographic characteristics that render
the residents in Metro Manila vulnerable to flooding. It also identifies and
ranks the key cities and barangays that are more vulnerable to flooding
due to the social characteristics of its residents.
Utilizing primary (Porio’s Japan Bank for International Cooperation
household survey data in 2011) as well as secondary (2010 National

57SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
Statistics Office Census of Population and Housing) data, this research
used principal component analysis to identify which socio-economic
characteristics are significantly correlated with social vulnerability
to flooding in Metro Manila. Furthermore, the study also constructed
social vulnerability indices for Metro Manila across city, barangay, and
household levels.
Examining the flood vulnerability of residents in Metro Manila,
the study argues that climate change-related effects put Metro Manila
cities at risk to flooding (Porio 2014). This research quantified social
vulnerability to flooding using principal component analysis, and then
ranked the cities according to their social vulnerability index scores. The
study concludes that components of social vulnerability vary across the
city, barangay, and household levels of analysis. Different components
of social vulnerability manifest at various levels of analysis, highlighting
the multi-dimensionality and interactivity of social vulnerability in Metro
Manila.
BACKGROUND AND CONTEXT
The Philippines is one of the most disaster-prone countries in the world.
(Loyzaga 2013; Porio 2011; Yumul, Servando, and Dimalanta 2011;
Zoleta-Nantes 2002). In terms of typhoons, the country gets an average
of 20 typhoons annually. It experiences several flooding incidences due
to its location relative to the paths of typhoons and the propagation of
monsoons (Ignacio and Henry 2013), and this causes heavy precipitation
and flooding to the capital city, Metro Manila (Porio 2011; Yumul,
Servando, and Dimalanta 2011). According to Porio (2014), the number
and scale of natural and human-induced disasters have increased in the
past decade. For example, in 2009, two major typhoons wreaked havoc in
Metro Manila, typhoon Ondoy (international name Ketsana) and Pepeng
(international name Parma), resulting in large numbers of affected people
and casualties.
The official death toll from the two major disasters combined was
956 persons, with 84 persons still missing and 736 injured (PDNA 2009).
Furthermore, the two typhoons resulted to a total of PHP 3.8 billion in
damages and PHP 24.8 billion in immediate losses to the agriculture,
fisheries and forestry sectors (Porio 2014). In September 2011, typhoons

5 8 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Pedring (international name Nesat) and Quiel (international name Nalgae)
claimed 90 lives, with an estimated amount of PHP 15 billion in damages
(Porio 2014). In December 2011, Typhoon Sendong (international name
Washi) brought heavy rainfall and flashfloods to Northern Mindanao and
Eastern Visayas (Porio 2014), leaving 1,002 dead and almost PHP 1 billion
in damages. Southwest monsoon rains (habagat) in 2012 (enhanced by
Typhoon Haikui) resulted to 112 deaths, affected 4.5 million persons, and
caused approximately PHP 3 billion in damages (NDRRMC 2012). In
2013, southwest monsoon rains (enhanced by Typhoon Maring) resulted
to 27 deaths, affected 3.1 million persons, and caused PHP 689 million
in damages. Finally, Typhoon Yolanda (international name Haiyan), one
of the world’s strongest storms in history that ripped through the middle
of the Philippine archipelago last November 8, 2013, caused widespread
damage and destruction in the country, affecting 16 million people, taking
6,300 lives, damaging 1,100 houses, and causing PHP 125 billion in total
damages. The occurrence of typhoons in the country has gone beyond
the regular typhoon season of June to November to throughout the year
(Porio 2012; 2014).
SOCIO-ECONOMIC AND DEMOGRAPHIC
CHARACTERISTICS OF METRO MANILA
Metro Manila has a land area of 636 square kilometers in semi-alluvial
plain formed by the sediment flows from the Meycauayan and Malabon-
Tullahan river basins in the North, the Pasig- Marikina river basin in the
East (Porio 2011; Bankoff 2003). According to Porio (2014), Manila and
the surrounding cities are prone to flooding alongside Marikina Valley
and along the coast of Laguna de Bay. Liongson (2000, cited in Bankoff
2003) described Metro Manila as “a vast drainage basin that experiences
frequent inundations from overflowing rivers and storm waters that
render the existing system of esteros (modified natural channels) and
canals constructed during the Spanish and American colonial periods
inadequate”.
Metro Manila has served as the main socio-economic and political
center of activities in the Philippines since the colonial period.
Consequently, this has greatly contributed to the increased vulnerability
of the residents, especially those from the urban poor communities, to

59SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
climate-related effects like typhoons and floods (Porio 2014). Its strategic
location by the Manila Bay and the mouth of the Pasig River accounts
for its growth and expansion in the last 30 years (Porio 2014). Manila
Bay provided an ideal port location for both foreign/local ships/boats to
dock in Manila Bay and near the mouth of the Pasig River to transport
of people and goods. Ports served as the transport hub for import and
export of goods starting with the Galleon trade between the Philippines
and Europe. Meanwhile, the Pasig River served as the main channel for
bringing people and goods from the hinterlands of Luzon to the city. This
all changed at the turn of the century with the increasing shift to road-
based transport and the decline of water-based transport along the Pasig
River and those docking in Manila Bay.
In 1970, Metro Manila was already 100% urbanized with a
population density of 6,200 persons per km2. By 1995, this population
density ballooned to about 14,930 persons per km2 (Mercado 1998).
As of 2013, the Philippines has a household population of 98,393,574,
with a population growth rate of 1.7%. Coupled with this population
growth is the fact that 41.7% of the population lives on less than $2.00
a day (World Bank 2013). Thus, poverty is widespread in the country,
in both urban and rural areas. Services and infrastructural developments
could hardly keep pace with the increasing population, and the growth
of both formal and informal settlements have been minimally regulated.
Buildings and infrastructure support are constantly being built on flood
and danger zones, compromising the quality of life in the metropolis
(Porio 2014). With an increasing population in Metro Manila, there is
also an increasing potential for loss of lives and properties (Loyzaga
2013; Porio 2011; Yumul, Servando, and Dimalanta 2011). Therefore,
investments in climate risk vulnerability assessments are important in
minimizing the losses from these calamities (Porio 2014; 2011).
CORRELATES OF SOCIAL VULNERABILITY
Within the social science literature, a considerable amount of research has
been focused on the indicators that either increase or decrease the impact
of particular hazard events on the population. Cutter et al. (2008) defines
these “indicators” as measures intended to represent a characteristic or
a parameter of a system of interest using a single value. By the 1960s,

6 0 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
social science researchers started using social indicators (Cutter et al.
2009). During 2000s and above, there has been more emphasis on the
development of environmental sustainability and vulnerability indicators
(Birkmann 2007; Polsky, Neff, and Yarnal 2007).
One of the most popular social indicators that affect the social
vulnerability of a community is poverty level (Few 2003; Adger 1999;
Chan and Parker 1996). Communities with a higher incidence of poverty
tend to be more vulnerable to environmental hazards. The poor tend
to occupy the more flood-prone environments, and they have fewer
resources upon which to draw to counteract the impacts of flooding (Porio
2011; Few 2003; Chan and Parker 1996). However, poverty is not the
only indicator of vulnerability to environmental hazards. Vulnerability
to the impacts of hazards has social, political, institutional, and cultural
dimensions as well (Few 2003; Cutter 2000; Pelling 1997).
Gender is another factor that affects social vulnerability. Studies have
also shown that women are more vulnerable than men (Loyzaga 2013;
Porio 2013; Neumayer and Plumper 2007; Peterson 1997). This effect
is strongest in countries with very low social and economic rights for
women. According to Neumayer and Plumper (2007), women suffer more
for two reasons: (1) women occupy a more tenuous position in society
prior to disasters, and (2) they have additional burden as caregivers to
children and the elderly. Hence, because women are more likely to be at
home and will risk their own lives to save their children and the elderly,
it makes escape to safety more difficult.
Both the young and the old may also be at a disadvantage in times of
calamities. Reviews on social vulnerability indicate that people who are
either very old or very young generally correlate with higher degrees of
mortality during floods. The findings conclude that people in both ends
of the age spectrum, because of their limited and/ or reduced physical
strength, are less likely to be able to withstand disasters, such as floods
(Tapsell, Penning-Rowsell, Tunstall, and Wilson 2002). They are also
more likely to experience health problems as well as experience a slower
recovery. The elderly may also experience distress at the possibility of
evacuation (Gladwin and Peacock 1997).
People with no land security and tenure are also more vulnerable
because they may usually lack the financial capacity to buy their own homes.

6 1SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
According to Porio (2014), a large portion of the population that does not
have security of tenure in housing, jobs, and livelihood sources heighten
the vulnerability of the community. Individuals with no land security tend
to not have adequate access to electrical, water, and sewerage services.
Furthermore, they may also lack access to information about disaster aid
during recovery (Cutter, Boruff, and Shirley 2003; Morrow 1999).
Individuals with lower education may usually be limited and
constrained in their ability to understand warning information and
access to recovery information. Furthermore, education is also linked to
socio-economic status, with higher educational attainment resulting to
higher monetary compensation and earnings, (Porio 2014; Steinführer
and Kuhlicke 2007; Cutter, Boruff, and Shirley 2003). Linked to an
individual’s education is his or her occupation. A number of occupations,
especially those involving resource extraction, may be severely affected
in times of natural hazards. Workers engaged in fishing, agriculture,
and other low-skilled service jobs may suffer as their incomes maybe
depleted when their livelihood sources are hit by hazards like floods and
tidal/storms surges. There is also a potential for employment loss after a
disaster (Porio 2011; Cutter, Boruff, and Shirley 2003; Hewitt 1997).
Finally, individuals who are totally dependent on social services
may also be limited and less able to respond effectively during disasters.
Thus, they may require more assistance and support, before, during and
post-disaster recovery period (Cutter, Boruff, and Shirley 2003; Morrow
1999). Differently-abled individuals may also be economically and
socially marginalized, and because of this, they may be mostly ignored
during recovery (Morrow 1999; Tobin and Ollenberger 1993). Table 1
presents a summary of a number of characteristics of social vulnerability,
as reported in a number of studies in the literature.
DATA SOURCES AND ANALYSIS
The dataset used in this research was twofold: primary data from the
household survey of Porio (2011), and secondary data obtained from the
2010 Census of Population and Housing (CPH) of the National Statistics
Office.
The primary data comes from the study of Porio (2011) for Japan Bank
International Cooperation (JBIC) titled Vulnerability, Adaptation, and

62 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Table 1: Socio-demographic Characteristics
and Vulnerability in the Literature
SOCIO -DEMOGRA PHIC
CHARA CTER ISTIC S
CHARA CTERISTI CS OF
HIGH VULNERABI LITY
SOURC E
Income Lower Income Porio (2014; 2011)
Loyzaga (2013)
Bankoff (2003)
Zoleta-Nantes (2002)
Age Old People Birkmann (2006)
Tapsell, Penning-Rowsell,
Tunstall, and Wilson (2002)
Very Young People Birkmann (2006)
Cola (1993)
Gender Female Porio (2014; 2013)
Porio (2011)
Neumayer and Plumper (2007)
Peterson (1997)
Education Lower education Steinführer and Kuhlicke (2007)
Tenure Renters Morrow (1999); Porio (2014)
Cutter, Boruff and Shirley (2003)
Social Dependence Differently-abled
population and
those who are
totally dependent
on social services
Morrow (1999)
Tobin and Ollenberger (1993)
Family Structure Single parents
Large number of
dependents
Cutter, Boruff, and Shirley, 2003
Morrow, 1999
Blaikie, 1994
Occupation People employed in
resource extraction
Cutter, Boruff, and Shirley (2003)
Hewitt (1997)
Infrastructure Loss of
infrastructure
Cutter, Boruff, and Shirley (2003)

6 3SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
Resilience to Flood and Climate Change-Related Risks Among Riverine
Communities in Metro Manila. The household survey was conducted
among 300 urban poor households in 14 communities located in the three
flood basins: 1) Pasig-Marikina river basin, (2) West Mangahan, and (3)
the CAMANAVA area (Caloocan, Malabon, Navotas, Valenzuela).
The secondary data comes from the census of the National Statistics
Office. The Census, also called as the “POPCEN”, is the source of
information on the size and distribution of the population as well as
information about the demographic, social, and economic characteristics.
The variables were selected based on the results of previous researches
and the availability of data. The 2010 census data set has a total of
92,097,978 individual person records, with a total housing of 21,745,707
individual person records.
METHODOLOGY
Principal Component Analysis (PCA) was used in this study to reduce
and group like variables into component groups for classification and
ease of further analysis. In this particular research, PCA was conducted at
three (3) different levels of analysis: (1) City Level, (2) Barangay Level
and (3) Household Level. Table 2 shows the socio-economic variables
used in this study.
Whereas the seventeen cities and municipalities were examined at the
city level, the community analysis was limited to Pasig barangays only
(for the barangay level analysis), and to CAMANAVA, Pasig, Marikina,
and Taguig households only at the household level (see Table 3). Pasig
was chosen at the barangay level, not only because the city regularly
experiences extreme rainfall events and flooding, but also because that
was the only data available at the barangay level. At the household
level, the survey conducted by Porio (2011) covered the households in
KAMANAVA, Pasig, Marikina, and Taguig cities only.
The Social Vulnerability Index (SoVI) methodology (adapted from
Cutter, Boruff, and Shirley 2003) was then used in this study. This
resulted to the following: (1) correlates of social vulnerability to Metro
Manila in the city, barangay, and household levels of analysis, and (2)
social vulnerability indices, which ranked places vulnerable to flooding.
The Social vulnerability indices constructed did not have a priori

6 4 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Table 2: Socio-Economic Variables and Vulnerability Concepts Addressed
CONCE PT OF
VULNE RABILITY AD APTED
FROM CUTTER, BORUFF,
AND S HIRLEY (2 003)
VULNE RABILITY INDICATORS
FROM THE NSO ( 2010) –
CITY AND BARAN GAY
LEVEL ANALYSIS
VULNE RABILITY INDICATORS
FROM PORIO (2 011) – HOU SE-
HOLD LEVEL ANA LYSIS
Age Percentage of population
below five years old
Percentage of population
below five years old
Percentage of population
with age of sixty-five and
above
Percentage of population
with age of sixty-five
and above
Gender Percentage Female Percentage Female
Special Needs Percentage of Population
Disabled
Percentage of Population
who are Occasionally
Sick
Percentage with
Difficulty in Walking
Percentage with
Difficulty in Seeing
Education Percentage of Population
that did not finish High
School
Percentage of Population
with Less than 10 Years
in School
Median Years in School
Family Structure Percentage of Population
Separated/ Divorced
Percentage Widower
Percentage Live-in
Percentage of Population
Widowed
Number of persons in
household
Percentage of Population
with No Social Networks
Socio-economic
Status
Proxy variables (tenure
and infrastructure) were
used.
Median Income
Percent Unemployed
Renters Percentage of Population
as Renters
Percentage of Renters
Percentage of Population
as Rent-Free with No
Consent

6 5SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
Table 3: Cities/Municipalities, Barangays, and Households subjected to PCA
CITY L EVEL
(SOUR CE: NSO 2010)
BARAN GAY LE VEL
(SOUR CE: NSO 2010)
HOUSE HOLD LEV EL
(SOUR CE: PORI O 2011)
Caloocan Pasig Caloocan
Malabon
Marikina
Navotas
Pasig
Taguig
Valenzuela
Las Piñas
Makati
Malabon
Mandaluyong
Manila
Marikina
Muntinlupa
Navotas
Parañaque
Pasay
Pasig
Pateros
Quezon City
San Juan
Taguig
Valenzuela
CONCE PT OF
VULNE RABILITY AD APTED
FROM CUTTER, BORUFF,
AND S HIRLEY (2 003)
VULNE RABILITY INDICATORS
FROM THE NSO ( 2010) –
CITY AND BARAN GAY
LEVEL ANALYSIS
VULNE RABILITY INDICATORS
FROM PORIO (2 011) – HOU SE-
HOLD LEVEL ANA LYSIS
Infrastructure Percentage of Population
with Roof Made of
Makeshift Materials
Percentage of
Households with Only
One Floor
Percentage of Population
with Outer Walls Made
of Makeshift Materials
Percentage of
Households with
Housing Materials Made
of Wood and Iron
Percentage of Population
with Dilapidated Houses
Percentage of Population
with No Electricity
Percentage of Population
with Tapped Water or
Deep Well
Percentage of Population
with No Septic Tank
Table 2 (cont’d)

6 6 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
assumptions about the weights of each factor in the overall sum. Thus,
each of the factors obtained was viewed as having an equal contribution
to the overall social vulnerability. The formula for index construction is
shown below:
Social Vulnerability Index = Component1 + Component2 + Component3 + …
The result was a univariate score that broadly represented the relative
levels of social vulnerability within the different cities, barangays and
households in Metro Manila. The index created had a range of values
from 0 to 1. Values close to 1 indicate higher social vulnerability to
flooding, whereas values close to 0 indicate lower social vulnerability to
flooding.
RESULTS AND DISCUSSION
City-Level Principal Component Analysis
At the city level, the Principal Component Analysis generated five
components accounting for 88.33% of the variance. The components are
shown in Table 4. They were named in terms of their representation of
social vulnerability, and were assigned a cardinal direction: positive (+) if
majority of the variables in the component increase vulnerability, negative
(-) if majority of the variables decrease vulnerability and the absolute
value if the component variables have a mixed impact on vulnerability.
The five components of social vulnerability at the city level are: (1)
Social Structure of the Community; (2) Disabilities and Difficulties; (3)
Housing Material; (4) Housing Condition; and (5) Land Tenure.
Social Structure of the Community. • The first factor identified is
the social structure of the community, and this is defined by the
following variables: percentage of children under 5 years old (0.914),
percentage women (0.894), percentage who have not completed high
school (0.805), and percentage of people over 65 years old (0.785).
The social structure factor explains 39.80% of the variance, and this
component has the highest factor loading.

67SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
Table 4: Findings of the City-Level Principal Component Analysis
COMPO NENT
NAME PERCE NT
VARIATION
EXPLA INED
DOMIN ANT VARIABLES COMPON ENT
LOADI NG
SCORE S (WITH
CARDI NAL
DIREC TIONS
ADJUS TED)
1 Social
Structure
of the
Community
39.80% Percent Under 5 Years Old +0.914
Percent Female +0.894
Percent of Population who
Have Not Completed High
school
+0.805
Percent of Population over
65 Years Old
+0.785
2 Disabilities
and
Difficulties
23.62% Percent with difficulty in
walking
+0.808
Percent with difficulty in
concentrating
+0.802
Percent with difficulty in
communicating
+0.709
Percent with difficulty in
seeing
+0.566
3 Housing
Material
12.30% Percent with Makeshift
Roof
+0.919
Percent with Makeshift Wall +0.772
4 Housing
Condition
7.01% Percent that Need Major
Repair
+0.726
Percent Dilapidated House +0.667
5 Housing &
Land Tenure
5.60% Percent Renter +0.941
Disabilities and Difficulties. • The second factor refers to disabilities
and difficulties, and includes the following variables: percentage
with difficulty in walking (0.808), percentage with difficulty in
concentrating (0.802), percentage with difficulty in communicating
(0.709), and percentage with difficulty in seeing (0.566). The second
factor explains 23.62% of the variation among the municipalities,
and has the second highest component loading.

6 8 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Housing Material and Housing Condition• . The third and fourth
factors refer to the type of housing material used and the physical
condition of the house. In this study, the following variables are
related: “percentage of houses with outer walls made of makeshift
materials” (0.772) and “percentage of houses with roofs made of
makeshift materials” (0.919) for the third factor, and “percentage
that need major repair” (0.726) and “percentage with dilapidated
houses” (0.667). The third factor (Housing Material) explains 12.30
% of the variance, while the fourth factor (Housing Condition)
explains 7.01 % of the variance.
Tenure Status. • The fifth factor identified is the tenure status, and in
this analysis, the significant variable is “percentage of households
who rent” (0.941). The fifth factor explains 5.60 % of the variation
among the municipalities.
A Social Vulnerability Index was then constructed for the 17 cities
and municipalities in Metro Manila (see Table 5).
Table 5: Social Vulnerability Index Scores for Metro Manila
SOCIA L VU LNERABILI TY RANKIN G CIT Y/ MUNICI PALI TY SOVI SCORE
1 (Most Vulnerable) Navotas 3.79
2 Malabon 3.04
3 Caloocan 2.16
4 Manila 1.74
5 Valenzuela 0.18
6 Parañaque 0.07
7 Marikina – 0.17
8 Las Piñas – 0.31
9 San Juan – 0.41
10 Muntinlupa – 0.58
11 Pateros – 0.58
12 Quezon City – 0.59
13 Mandaluyong – 0.73
14 Pasig – 1.49
15 Pasay – 1.94
16 Makati – 1.94
17 (Least Vulnerable) Taguig – 2.25

69SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
The top five cities in Metro Manila with the highest social vulnerability
include: Navotas (3.79), Malabon (3.04), Caloocan (2.16), Manila
(1.74), and Valenzuela (0.18). Four of the five highly vulnerable cities
to flooding are located in the KAMANAVA area, which regularly suffer
from extreme rainfall events, typhoons, and floods. According to Muto et
al (2010), KAMANAVA area is low and flat with elevations ranging from
around sea level to 2 meters above sea level. Furthermore, it was made
up of widely spread lagoons used as fishponds during the 1960s. The
geo-physical context of vulnerability of the KAMANAVA area, seemed
to have heightened the social vulnerability of their residents, and vice-
versa.
Barangay-Level Principal Component Analysis
At the barangay level, the PCA generated four (4) components
accounting for 76.35% of the variance. Table 6 shows the results for
the barangays in Pasig City. Not surprisingly, the statistically significant
indicators of social vulnerability at the barangay level were similar to and
consistent with the indicators at the city level (Table 4).
The first component, social structure of the community, comprised
of 33.15% of the variance. There were five significant contributors to
social vulnerability: percent female (0.899), percent who did not finish
high school (0.875), percent below 5 years old (0.838), percent above 65
years old (0.779), and percent separated (0.656).
The second component, disabilities and difficulties, comprised of
19.89% of the variance. There were four significant factors that contribute
to social vulnerability: percent with difficulty hearing (0.908), percent
with difficulty walking (0.898), percent with difficulty communicating
(0.758), and percent with difficulty seeing (0.689).
The third component, housing materials and condition, comprised
of 10.23% of the variance. There were two significant contributors to
social vulnerability: percent with makeshift walls (0.771), and percent
with dilapidated houses (0.760).
Finally, the last component, land tenure, comprised of 13.08% of
the variance. There was one significant factor that contributes to social
vulnerability: percent of people who rent (0.709).

70 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Table 6: Findings of the Barangay-Level Principal Component Analysis
COMPO NENT
NUMBE R
NAME
% OF
VARIANCE
EXPLA INED
VARIABLES
COMPO NENT
LOADI NG
1 Social
Structure of the
Community
33.15 Percentage Female +0.899
Percentage who did not
finish high school
+0.875
Percentage 5 years old and
below
+0.838
Percentage 65 and above +0.779
Percentage separated/
divorced
+0.656
2 Disabilities and
Difficulties
19.89 Percentage with difficulty
hearing
+0.908
Percentage with difficulty
walking
+0.898
Percentage with difficulty
communicating
+0.758
Percentage with difficulty
seeing
+0.689
3 Housing
Materials and
Condition
10.23 Percentage with makeshift
walls
+0.771
Percentage dilapidated
houses
+0.760
4 Land Tenure 13.082 Percentage people renting +0.709
Household-level Principal Component Analysis
At the household level, five (5) components accounting for 80.01%
of the variance were extracted. The five components that contribute to
social vulnerability at the household level are: (1) Education; (2) Social
structure and network; (3) Income and access to services; (4) Housing
condition; and (5) Land tenure (see Table 7).
Education• . The first factor identified is the education of the
community, and this is defined by the following variables: Median

7 1SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
Table 7: Findings of the Household Level Principal Component Analysis
COMPO NENT
NAME
PERCE NT
VARIATION
EXPLA INED
DOMIN ANT VARIABLES
COMPO NENT
LOADI NG
SCORE S (WITH
CARDI NAL DIREC-
TIONS ADJ USTED)
1 Education 28.88% Median Years in School – 0.910
Percentage of Households
who have not finished High
School
+0.907
Percentage of Students
with Less than 10 Years in
School
+0.769
2 Social Structure
and Network
21.23% Number of Persons per
Household
+ 0.802
Percentage Female + 0.736
Percentage of Population
with No Social Networks
+ 0.728
3 Income and
Access to
Services
13.51% Percentage of Households
with No Septic Tank Toilet
+0.797
Percentage of Households
with No Electricity
+ 0.768
Median Monthly Income – 0.623
4 Housing
Condition
9.82% Percentage of Households
with Only One Floor
+0.836
Percentage of Population
with Housing Material of
Wood and Iron
+0.738
5 Land &
Housing Tenure
6.57% Percentage of Renters +0.893
Years in School (-0.910), Percentage of Households who have not
finished High School (0.907), and Percentage of Students with Less
than 10 Years in School (0.769). This factor accounted for 28.88%
of the variance.
Social Structure and Network.• The second factor refers to the social
structure and network of the household population, which includes

7 2 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
the following variables: Number of Persons per Household (0.802),
Percentage Female (0.736), and Percentage of Population with No
Social Networks (0.728). The second factor accounted for 21.23% of
the variance.
Income and Access to Basic Services• . The third factor refers to type
of housing material used, and in this study, the following variables
were related: Percentage of Households with No Septic Tank Toilet
(0.797), Percentage of Households with No Electricity (0.768), and
Median Monthly Income (-0.623). The third factor explained 13.51%
of the variance.
Housing Condition• . The fourth factor identified is the housing
condition, and in this analysis, the significant variable related to
this are the following: Percentage of Households with Only One
Floor (0.836), and Percentage of Population with Housing Material
of Wood and Iron (0.738). This factor accounted for 9.82% of the
variance.
Housing and Land Tenure• . The fifth factor refers to the tenure status
of the house, and in this study, the variable related to this factor was
“percentage of households with tenure status as rented” (0.893). The
fifth factor accounted for 6.57% of the variance.
Note that there were two new components of social vulnerability
that appeared to be statistically significant at the household level: social
networks of the people (component 2) and their access to basic services
(component 3). Thus, the measure of social vulnerability was refined by
expanding the analysis with the inclusion of household level data.
Based on the findings of the PCA at the household level, a Social
Vulnerability Index was once again constructed. Table 8 shows the Social
Vulnerability Index for the 14 barangays in the survey of Porio (2011).
The barangays with the highest social vulnerability are the following:
Napindan, Taguig (3.54), Calzada, Taguig (3.49), Ibayo Tipaz, Taguig
(2.94), San Agustin, Malabon (2.78), and Longos, Malabon (2.78).
These findings are consistent with the results of the study of Porio
(2011) in her article entitled Vulnerability, Adaptation, and Resilience to
Flood and Climate Change-Related Risks Among Riverine Communities
in Metro Manila. Porio found a strong interaction between the

7 3SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
environmental-ecological vulnerability of the riverine communities with
the social vulnerability of the urban poor in these areas (Porio 2011). Of
the three flood plains in her study, the residents in the West Manggahan
and CAMANAVA flood plains (who have less income and education) also
tend to be more vulnerable and thus suffer more loses and inconveniences
due to flooding.
In summary, the social structure of the community, disabilities,
housing materials and conditions, and housing/land tenure are the socio-
economic characteristics that are significantly correlated with social
vulnerability to flooding in Metro Manila at both the city and barangay
levels, and they are consistent with the findings in the vulnerability
literature. Studies show that the following may be more vulnerable to
flooding:
The poor • as they may have little or no savings, few income or
production options, and limited resources (Porio 2011; Bankoff
2003; Zoleta-Nantes 2002)
• The very young and the very old because they might physically be
at a disadvantage, as well as the possibility for lack of material and
Table 8: Social Vulnerability Index Scores for the 14 Ba-
rangays in the Household Survey of Porio (2011)
SOCIA L VU LNERABILI TY
RANKI NG
BARAN GAY SOCI AL VULNERAB ILITY
INDEX SCORE
1 (Most Vulnerable) Napindan, Taguig 3.54
2 Calzada, Taguig 3.49
3 Ibayo Tipaz, Taguig 2.94
4 San Agustin, Malabon 2.78
5 Longos, Malabon 2.78
6 Bangculasi, Navotas 2.24
7 Bagumbayan South, Navotas 1.31
8 West Navotas 1.08
9 Bgy. 28, Caloocan 0.45
10 Bagong Ilog, Pasig – 0.51
11 San Joaquin, Pasig – 1.45
12 Rosario, Pasig – 1.49
13 Tumana, Marikina – 2.66
14 (Least Vulnerable) Ugong, Pasig – 2.83

74 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
economic support (Birkmann 2007; Rygel, O’Sullivan, and Yarnal
2005; Zoleta-Nantes 2002)
The women • as they may also carry the burden of responsibility
towards their children and the elders aside from their own safety
(Porio 2011; Neumayer and Plumper 2007)
• Those with little education as they may be constrained in their ability
to understand warning information and access to recovery information
(Steinführer and Kuhlicke 2007; Cutter, Boruff and Shirley 2003;).
The mentally or physically disabled or impaired• as their mobility is
limited and they may be unable to effectively respond to, and may
require the assistance of others during disasters (Rygel, O’Sullivan,
and Yarnal 2005).
Those living in temporary/makeshift houses in informal settlements• ,
as they are prone to environmental hazards such as typhoons and
earthquakes (Porio 2011)
Those without security of tenure• , as they usually do not have the
financial capability to buy their own homes, and they often lack
access to information about financial aid during disasters (Porio 2014
and 2011; Cutter, Boruff, and Shirley 2003; )
Finally, at the household level, two additional factors were found to
be significant contributors to social vulnerability: social networks and
access to basic services. In the literature, people who do not have a wide
network of relatives, neighbors, and friends, as well as those who cannot
access aid from formal institutions reflect the thinness of the social capital
they can rely on in times of calamities (Porio 2015; 2011).
This gives them limited options and thus makes them less able to act
on adaptation (Flanagan Gregory, Hallisey, Heitgerd and Lewis 2011).
Furthermore, people who do not have adequate access to basic services
such as electricity, water, sewerage, and drainage systems do not only
regularly suffer from floods, but they are also prone to sicknesses and
diseases (Porio 2014; 2011). Thus, limitations in access to basic services
also constrain options for strengthening communities to manage climate
risks (CARE 2011).

7 5SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
CONCLUSION
The study highlighted a number of important dimensions of social
vulnerability to flooding in Metro Manila. First, one indicator alone,
such as percentage of the elderly, cannot be used as a measure for social
vulnerability. Instead, a combination of all the indicators is needed to
accurately calculate social vulnerability. Principal Component Analysis
might be used as a tool in measuring social vulnerability. Second, there
are different aspects of vulnerability that can manifest at different levels
of intervention. Whereas housing materials/ conditions and land tenure
were the consistent components of social vulnerability at the city and
barangay levels, social networks, and access to basic services proved to
be significant at the household level. Third, the methodologies to be used
in measuring social vulnerability should be adapted to best reflect the
local contexts of the community. The methodology of Cutter, Boruff, and
Shirley (2003) is applicable to Metro Manila, but only to a certain extent
because of the restricted data set in the Philippine Census Data. Hence,
proxy variables were used.
Finally, since this is part of an on-going research, there are still several
steps that need to be done to come up with a more refined and enhanced
measure of social vulnerability to flooding in Metro Manila. One is the
possibility of including more indicators of social vulnerability into the
analysis. Other factors such as religion, ethnicity, languages spoken,
etc. might come up to be significant indicators of social vulnerability.
Furthermore, incorporating a number of ecological-environmental
variables into the principal component analysis could also enhance the
measure of social vulnerability. Finally, an assessment of how social
vulnerability changes over time will also be very relevant, given the
dynamic nature of social vulnerability in the literature.
ACKNOWLEDGEMENTS
The authors would like to thank the following:
The Manila Observatory; •
The Ateneo de Manila University, especially the Department of Sociology and •
Anthropology;

76 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
The International Research Initiative on Adaptation to Climate Change •
(IRIACC), managed by the International Development Research Center (IDRC),
Canadian Institute of Health Research, Natural Sciences and Engineering
Research Council, and the Social Science and Humanities Research Council;
The National Statistics Office (NSO), most especially their administrator •
Carmelita N. Ericta, as well as the staff of the Databank and Information
Services Division (DISD) for all the assistance given to the authors.
REFERENCES
Adger, W. Neil. 1999. “Social Vulnerability to Climate Change and Extremes in
Coastal Vietnam.” World Development 27(2):249-269.
Adger, W. Neil. 2006. “Vulnerability.” Global Environmental Change 16(3):268-281.
Bankoff, Greg. 2003. “Constructing Vulnerability: The Historical, Natural, and Social
Generation of Flooding in Metropolitan Manila.” Disasters 27(3):224-238.
Birkmann, Jörn. 2007. “Risk and Vulnerability Indicators at Different Scales:
Applicability, Usefulness and Policy Implications.” Environmental Hazards
7(1):20-31.
Birkmann, Jörn and Ben Wisner. 2006. “Measuring the Unmeasurable.” UNU-EHS
SOURCE, 5. Retrieved February 23, 2014 (http://www.ihdp.unu.edu/file/
get/3962.pdf).
Birkmann, Jörn. 2006. Measuring Vulnerability to Natural Hazards: Towards
Disaster Resilient Societies. New York: United Nations Publications.
Blaikie, Piers, Terry Cannon, Ian Davis, and Ben Wisner. 1994. At Risk: Natural
Hazards, People’s Vulnerability, and Disasters. London: Routledge.
CARE, 2011. “Understanding Vulnerability to Climate Change: Insights from
Application of CARE’s Climate Vulnerability and Capacity Analysis
(CVCA) Methodology.” CARE International Poverty, Environment and
Climate Change Network (PECCN).
Cola, Raoul M. 1993. “The Needs of Children Following a Disaster: The 1990
Earthquake in the Philippines.” Disasters 17(3):248-253.
Cutter, Susan L. 1996. “Vulnerability to Environmental Hazards.” Progress in
Human Geography 20(4):529-39.
Cutter, Susan L., Bryan J. Boruff, and W. Lynn Shirley. 2003. “Social Vulnerability
to Environmental Hazards.” Social Science Quarterly 84(1):242-261.

7 7SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
Cutter, Susan L., Lindsey Barnes, Melissa Berry, Christopher Burton, Elijah
Evans, Eric Tate, and Jennifer Webb. 2008. “A Place-Based Model for
Understanding Community Resilience to Natural Disasters.” Global
Environmental Change 18(4):598-606.
Cutter, Susan L., Christopher T. Emrich, Jennifer J. Webb, and Daniel Morath. 2009.
Social Vulnerability to Climate Variability Hazards: A Review of Literature.
Columbia: University of South Carolina– Hazards & Vulnerability Research
Institute.
Downing, Thomas E., Alexander A. Olsthoorn, and Richard S.J. Tol. 1999. Climate,
Change and Risk. London: Routledge.
Downing, Thomas E., Ruth Butterfield, Stewart Cohen, Saleemul Hug, Richard
Moss, A. Rahman, Youba Sokona, and Linda Stephen. 2001. Vulnerability
Indices: Climate Change Impacts and Adaptation. UNEP Policy Series.
UNEP, Nairobi.
Few, Roger. 2003. “Flooding, Vulnerability, and Coping Strategies: Local Responses
to a Global Threat.” Progress in Development Studies 3(1):43-58.
Flanagan, Barry, Edward Gregory, Elaine Hallisey, Janet Heitgerd, and Brian Lewis.
2011. “A Social Vulnerability Index for Disaster Management.” Journal of
Homeland Security and Emergency Management 8(1):Article 3.
Gladwin, Hugh and Walter G. Peacock. 1997. “Warning and Evacuation: A Night for
Hard Houses.” Pp. 52-72 in Hurricane Andrew: Ethnicity, Gender, and the
Sociology of Disaster, edited by Walter G. Peacock, Hugh Gladwin and Betty
H. Morrow. Miami: International Hurricane Center, Laboratory for Social
and Behavioral Research.
Hewitt, Kenneth. 1997. Regions of Risk: A Geographical Introduction to Disasters.
Essex: Longman.
Ignacio, Jose Andres F. and Sabin Henry. 2013. “Assessing the Vulnerability of
Populations at High Risk to Coastal River Flooding in the Philippines.”
Pp. 78-92 in From Social Vulnerability to Resilience: Measuring Progress
toward Disaster Risk Reduction, edited by Susan L. Cutter and Cosmin
Corendea. Bonn: United Nations University Institute for Environment and
Human Security.
Liongson, Leonardo Q., 2000. “The Esteros of Manila: Urban Drainage a Century
Since.” Pp. 1-16 in Pressures of Urbanization: Flood Control and Drainage
in Metro Manila, edited by Leonardo Q. Liongson, Guillermo Q. Tabios, and

78 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Peter P.M. Castro. Quezon City: Center for Integrative and Development
Studies, University of the Philippines Diliman.
Loyzaga, Antonia Y. 2013. Coastal Mega-cities at Risk. Conference on Peri-Urban
Development and Water Security. Crowne Hotel, June 18-21, 2013.
Kathmadu, Nepal.
Morrow, Brian H. 1999. ‘‘Identifying and Mapping Community Vulnerability.’’
Disasters 23(1):11-18.
Muto, Megumi, K. Morishita, and L. Syson. 2010. Impacts of Climate Change upon
Asian Coastal Areas: The Case of Metro Manila. JICA Manila Report as
JICA Research Institute Report.
National Disaster Risk Reduction and Management Council. 2012. Final Report on
the Effects of Southwest Monsoon and Emergency Response Management.
Quezon City: NDRMMC.
Neumayer, Eric and Thomas Plumper. 2007. “The Gendered Nature of Natural
Disasters: The Impact of Catastrophic Events on the Gender Gap in
Life Expectancy, 1981-2002.” Annals of the Association of American
Geographers 97(3):551-566.
Oliver-Smith, Anthony. 2008. “Sea Level Rise and the Vulnerability of
Coastal Peoples Responding to the Local Challenges of Global
Climate Change in the 21st Century.” Working Paper. Retrieved
March 15, 2014 (http://humansecuritygateway.com/documents/UNU_
impactenvironmentaldegradation.pdf).
Pelling, Mark. 1997. “What Determines Vulnerability to Floods: A Case Study in
Georgetown, Guyana.” Environment and Urbanization 9(1):203-26.
Peterson, Kristina. 1997. “From the Field: Gender Issues in Disaster Response and
Recovery.” Natural Hazards Observer 21(5):3-4.
Polsky, Colin, Rob Neff, and Brent Yarnal. 2007. “Building Comparable Global
Change Vulnerability Assessments: The Vulnerability Scoping Diagram.”
Global Environmental Change 17(3-4):472-485.
Porio, Emma. 2015. ”Not Business as Usual II: Science, Community and
Practitioners Conversations on Yolanda and the Challenge of the New
Normal”. Presented at Social Capital/Trust Networks, Post-Disaster
Recovery and Community Resilience Paper, Manila Observatory, and the
Ateneo de Manila University in partnership with Christian Aid and Aksyon
Klima-Philippines, La Breza Hotel, August 25, 2015.

7 9SEE-PORIOt4PDJBM7VMOFSBCJMJUZJO.FUSP.BOJMB'MPPEJOH
Porio, Emma. 2014. “Climate Change Vulnerability and Adaptation in Metro
Manila: Challenging Governance and Human Security Needs of Urban Poor
Communities.” Asian Journal of Social Science 42:75-102.
Porio, Emma. 2013. “Life Along Manila’s Flooding Rivers.” Pp. 256-270 in The
Environments of the Poor in Southeast Asia, East Asia, and the Pacific,
edited by Aris Ananta and Armin Bauer. Singapore: ISEAS Press.
Porio, Emma. 2012. “Enhancing Climate Change Adaptation Through Integration
to Long-Term Development Plans and Disaster Management: The Case of
Metro Manila.” Report submitted to the Asia Pacific Network for Global
Change, Tokyo, Japan.
Porio, Emma. 2011. “Vulnerability, Adaptation, and Resilience to Floods and Climate
Change-Related Risks Among Marginal, Riverine Communities in Metro
Manila.” Asian Journal of Social Science 39:425- 445.
Post-Disaster Needs Assessment Main Report Vol. 1. 2009. Retrieved February 23,
2014 (http://www.pdf.ph/downloads/PDNA/PDNA%20Main%20Report%20
%28Vol%201%29.pdf).
Rygel, Lisa, David O’Sullivan, and Brent Yarnal. 2005. “A Method for Constructing
a Social Vulnerability Index, Submitted to Mitigation and Adaptation
Strategies for Global Change”. Mitigation and Adaptation Strategies For
Global Change 11(3):741-764.
Steinführer, Annett and Kuhlicke Christian. 2007. “Social Vulnerability and the 2002
Flood”. Country Report Germany (Mulde River), Floodsite Report Number
T11-07-08. Leipzig: FZ Centre for Environmental Research.
Tapsell, Sue, Edmund Penning-Rowsell, Sylvia M. Tunstall, and Thomas L.
Wilson. 2002. “Vulnerability to Flooding: Health and Social Dimensions.”
Philosophical Transactions of the Royal Society A(360):1511-1525.
The World Bank. 2013. Philippines. Retrieved February 23, 2014 (http://data.
worldbank.org/country/philippines).
Tierney, Kathleen J., Michael K. Lindell, and Ronald W. Perry. 2001. Facing the
Unexpected: Disaster Preparedness and Response in the United States.
Washington: Joseph Henry Press.
Tobin, Graham A. and Jane C. Ollenburger. 1993. Natural Hazards and the
Elderly. Boulder: University of Colorado, Natural Hazards Research and
Applications Information Center.
United Nations University-Environment and Human Security. 2014. World Risk
Report 2014. Available (www.worldriskreport.org).

8 0 1IJMJQQJOF4PDJPMPHJDBM3FWJFXt7PMt4QFDJBM*TTVF
Yumul, Graciano, Nathaniel Cruz, Nathaniel Servando, and Carla Dimalanta. 2011.
“Extreme Weather Events and Related Disasters in the Philippines, 2004-
2008: A Sign of What Climate Change Will Mean?” Disasters 35(2):362-
382.
Zoleta-Nantes, Doracie. 2002. “Differential Impacts of Flood Hazards Among the
Street Children, Urban Poor, the Urban Poor, and Residents of Wealthy
Neighbourhoods in Metro Manila, Philippines.” Mitigation and Adaptation
Strategies for Global Change 7(3):239-266.