Content uploaded by Edmund C. Penning-Rowsell
Author content
All content in this area was uploaded by Edmund C. Penning-Rowsell on Jun 20, 2014
Content may be subject to copyright.
doi: 10.1098/rsta.2002.1013
, 1511-1525360 2002 Phil. Trans. R. Soc. Lond. A
S. M. Tapsell, E. C. Penning-Rowsell, S. M. Tunstall and T. L. Wilson
Vulnerability to flooding: health and social dimensions
References
http://rsta.royalsocietypublishing.org/content/360/1796/1511#related-urls
Article cited in:
Email alerting service
herecorner of the article or click
Receive free email alerts when new articles cite this article - sign up in the box at the top right-hand
http://rsta.royalsocietypublishing.org/subscriptions go to: Phil. Trans. R. Soc. Lond. ATo subscribe to
This journal is © 2002 The Royal Society
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
10.1098/rsta.2002.1013
Vulnerability to flooding: health
and social dimensions
By S. M. Tapsell, E. C. Penning-Rowsell,
S. M. Tunstall and T. L. Wilson
Flood Hazard Research Centre, Middlesex University,
Queensway, Enfield, Middlesex EN3 4SF, UK
Published online 24 May 2002
This paper presents research results on the impacts that floods can have on the
people affected, thus complementing the existing data on the monetary losses liable
to occur in flood events. Both datasets should be used when deciding on investment in
flood defence measures. We report on research on the vulnerability of flood-affected
communities to adverse health effects, and the development of an index of community
vulnerability based on extensive focus-group research and secondary-source census
data.
Keywords: floods; impacts; health; community; vulnerability
1. Introduction
One of the key issues in managing flood risk with the threat of climate change is to
determine the kind, and optimize the levels, of investment in flood defence measures.
The prime function of this investment—and thereby managing flood risk and flood
plains—is to protect people, property and precious environments from the damage
and disruption that floods can bring. The prospect of an increasing incidence of
flooding in a changing climatic regime does not alter this imperative.
Protecting property from flooding is catered for in investment decision making
through its potential flood damage being counted within the fairly rigorous benefit–
cost test that the UK Treasury insists is applied (Penning-Rowsell & Green 2000).
Environments are protected—now sometimes irrespective of cost—by courtesy of
DEFRA rules and European legislation creating SPAs, SSSIs, etc. But remaining
relatively ill considered are the people affected by floods and the impacts that they
suffer.
In this respect we need to understand better the ‘social’ effects of floods: those
caused by the disruption of people and communities that do not or cannot carry
a monetary price tag (Green et al. 1994). Floods can cause health impacts which
are enduring, including the stress and trauma created months or years afterwards
whenever floods again appear to threaten. These can be severe. Loss of treasured
possessions in floods can be ‘heartbreaking’, and much more significant than finan-
cial losses, which are now commonly recovered through household insurance policies
(Penning-Rowsell & Green 2000).
One contribution of 18 to a Discussion Meeting ‘Flood risk in a changing climate’.
Phil. Trans. R. Soc. Lond. A (2002) 360, 1511–1525
1511
c
2002 The Royal Society
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
1512 S. M. Tapsell and others
What we cannot easily do is measure these ‘intangible’ effects in monetary terms,
to build into the benefit–cost tests, and indeed we would not advocate that we try
to do this. What we can do is to record these impacts in much more detail than has
hitherto occurred, so that a proper balance can be struck between these effects and
the more easily measurable money losses. We can also attempt to develop predictive
models of these social effects, so that we can also plan to alleviate these in the future
with our policies, plans and schemes—just as we plan to reduce flood damage—rather
than just record them in our post-flood surveys.
This paper therefore contributes to the process of assessing the ‘people dimension’
that is now largely ignored in decisions about flood defence: we have methods to
gauge the economic gains and losses from such investment decisions but not the
degraded life quality that follows floods for many of their victims. We also report on
research that seeks to predict the ‘social impacts’ of floods so that we can alleviate
them more purposefully. Both will be increasingly important if climate change does
indeed bring greater flood threat.
2. Health and community flood impacts: research in
northeast England following the June 2000 floods
(a) Research focus and methods
Between 3 and 5 June 2000, parts of northeast England experienced flooding which
was unusually severe and extensive for the time of year (Environment Agency 2000).
On the evening of 3 June, high-intensity localized rain fell in a number of areas,
flooding over 1000 properties. Three of the communities affected were Todmorden in
Calderdale, Lancashire, and West Auckland and South Church, in County Durham.
In Todmorden the Upper Calder River flooded approximately 500 properties, while
in West Auckland and South Church some 430 were flooded by the River Gaunless
in an event having a return period of ca. 1:100 years (Environment Agency 2000).
In Todmorden a full flood-warning service was in place, while in West Auckland
and South Church no such service then existed. However, the ‘flashy’ nature of the
catchments, in addition to other complicating factors such as the collapse in Todmor-
den of two flood walls, meant that warnings were not issued in any of the locations
prior to flood waters entering properties. Flooding in all three locations occurred at
night, thus making response by emergency services and local residents more problem-
atic. The flooding, therefore, had considerable impact upon individuals, households
and communities in the three locations.
Our model of flood impacts (figure 1) is based on more than a decade of research
on the social dimensions of floods (Penning-Rowsell & Green 2000). It sees these
impacts as the net effect of the threat, the mediating influences that moderate that
threat for the affected population, and the support capacity in households, com-
munities and indeed the nation that helps to promote resilience in that population
and the capacity to recover from the threat, the event and its effects. Our methods
involved a qualitative approach using focus groups to match previous research in the
Thames region (Tapsell et al . 1999). A focus group is a carefully selected group of
people who meet to discuss particular questions raised by a moderator. Focus-group
discussions enable these people to share their views and feelings about an issue. In
the past such qualitative research was not valued because it was not seen as facilitat-
ing generalization and theory building. However, these methods are now seen to have
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
Vulnerability to flooding: health and social dimensions 1513
mediating
influences
threat
impacts
population
support
capacity
Figure 1. The process model of vulnerability to flooding: threats,
impacts and mediating influences (courtesy of Colin Green).
a key role to play in explaining and interpreting people’s lives, actions, perceptions,
fears and feelings, particularly when researching sensitive issues such as flooding.
The reporting of qualitative data is suggested by Miles & Huberman (1994, p. 299)
as possibly being ‘one of the most fertile fields going [as] there are no fixed formats’.
A total of six focus-group meetings were carried out in September and October
2000, three to four months after the flooding. Four meetings were held in Todmorden
and one each in West Auckland and South Church. All the focus-group sessions
were tape recorded with the participants’ permission and later fully transcribed. An
observer was present, along with the moderator, to take notes and observe responses.
Only a summary of the full results can be presented here. Table 1 gives information
on the key elements of the range of results obtained, and this paper concentrates
hereafter on health effects. For full details see Tapsell & Tunstall (2001).
(b) Physical health effects†
Previous research has indicated that natural disasters such as flooding can impact
upon people’s health in a number of ways (Parker et al. 1983; Green et al. 1985a;
Tapsell et al. 1999; Ohl & Tapsell 2000), good health being defined as complete phys-
ical, mental and social well-being (World Health Organization 1948). The physical
health effects reported by focus-group participants after the June 2000 flooding are
shown in table 2. These largely correspond to effects reported by respondents in the
earlier Thames region study (Tapsell et al. 1999), and many are associated with the
trauma of the flooding and living subsequently for long periods in damp and dirty
conditions.
Many people reported suffering from diarrhoea and upset stomachs following the
floods, while others spoke of other types of infection. The close proximity of people
living in cramped conditions in their homes following the flood also meant that
some of these adverse health effects were passed from person to person within the
household. Several people spoke of feeling generally unwell, although they could
not give any specific symptoms. There was also the suggestion that the full health
impacts of the flood have yet to reveal themselves. This was also mentioned by people
† Detailed quotations from the focus-group members to support the summaries in this section are
available in Tapsell & Tunstall (2001).
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
1514 S. M. Tapsell and others
Table 1. Summary results from the focus-group meeting surveys in the northeast of England
following the spring 2000 floods (focus-group meetings held in October/November 2000)
Key summary points: attitudes, stresses and behaviour before the flooding
Before the flooding there was little awareness or expectation of the risk of serious
flooding in Todmorden and West Auckland and no awareness in South Church.
Although some people had experienced past flooding in their properties in Todmorden
and West Auckland, in only a few of these cases was this serious.
The majority of people were not prepared to cope with the flooding.
People generally felt that the risk of flooding should have been made clearer by the
Environment Agency.
Key summary points: attitudes, stresses and behaviour during the flooding
During the flood people had been shocked at what was happening, and at the power,
speed and depth of the flood waters.
The risk to life from the flooding was highlighted by many people.
Most people had received informal warnings which had allowed some to save a few
possessions; apart from this there was little people could do.
Some help was received from emergency services, friends and neighbours, but many
people had to help themselves.
Various authorities, including the Environment Agency, were criticized for their lack of
support.
Key summary points: attitudes, stresses and behaviour after the flooding
Damage to property and losses from the flood were extensive. The most important
losses were irreplaceable personal items and memorabilia.
For those who were evacuated from their homes the experience was stressful and several
people had still not returned to their properties. Little rental accommodation was
available locally.
Those who did not evacuate faced months of living in damp and dusty conditions and
the prospect of being surrounded by empty properties. Disruption to daily life was
therefore great among both groups.
Taking time off work to recover from the flood had caused problems for people, not
least through the loss of income, but for a few people going to work offered some respite
from the flood recovery process.
Local authorities were generally criticized over what was perceived as insufficient
support during the recovery process.
The main forms of support required were suggested as being ‘manpower’, advice and
counselling. Voluntary support was generally well received.
Key problems were experienced with loss adjusters and insurance companies,
particularly regarding differing levels of service offered. Those without insurance faced
additional problems.
Builders and contractors repairing properties were also heavily criticized for their poor
standards of service, unreliability and unpleasant attitudes.
Strong feelings were expressed of having to ‘fight’ for any advice and assistance in the
recovery process. The effects of this had significant implications for people’s health and
well-being.
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
Vulnerability to flooding: health and social dimensions 1515
Table 1. (Cont.)
Key summary points: health effects
Many people had suffered from physical health problems since the flooding, often
associated with coming into contact with contaminated flood waters, living in damp
properties, etc.
Concern was expressed over the health risks from contaminated flood waters and the
associated lack of, or conflicting, advice given by various authorities on these health
issues.
The majority of participants admitted to feeling extremely ‘stressed’ by the flooding
and recovery process and some were displaying signs of common mental disorders
associated with experiencing a traumatic event.
Anxiety during rainfall was common among focus-group participants since the flooding,
and many had adjusted their behaviour (e.g. by regularly monitoring the river levels)
due to the fear of possible future flooding.
The most devastating aspects of the flooding were largely said to be financial, no time
for ‘living’, the loss of ‘everything people had worked hard for’, and feelings of
helplessness.
Key summary points: security and community cohesion
Issues related to future security concerned the fear of future flooding and the loss of
security people now feel in their homes—no longer a safe refuge.
Homes no longer have the same meaning for people as they did before the flooding.
Many people felt the need to know how to protect their homes from any future flooding.
The media were perceived as being intrusive and insensitive following the flooding.
Significant adverse effects were demonstrated by the disruption to, and deterioration of,
community life, particularly in West Auckland and South Church.
Key summary points: attitudes to the authorities
The general feeling was that the various authorities had responded poorly following the
flooding, particularly the local Council.
People had little confidence in the authorities to predict future flooding, provide timely
warnings or provide support.
The Environment Agency was the only agency perceived to be taking some action,
although these actions were frequently seen as being too little and too late.
The perception of focus-group participants was that the authorities, including the
Environment Agency, had no real commitment to flood prevention.
Various factors were seen to have caused or exacerbated the flooding—some of these
were accurate and some were not.
following the 1998 flooding, when it was only after people’s homes had been put back
in order that the full extent of what had happened was appreciated (Tapsell 2000).
This further indicates that the effects of flooding on people’s health and general
well-being can continue for many months after the actual flood event.
The research reported here shows that people’s health before flooding can be a
factor affecting the extent of health impacts experienced by flood victims. A num-
ber of preexisting health problems were highlighted in the focus groups by flooded
individuals, either for themselves or for close members of their families. For exam-
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
1516 S. M. Tapsell and others
Table 1. (Cont.)
Key summary points: vulnerable groups
Certain groups within the communities affected by the flooding were seen to have
suffered more pronounced effects than others. These groups were thought to need
particular support and consideration.
The flood was perceived to have had significant impacts on children, women, the elderly
and disabled, both physically and psychologically.
Children were thought to have been ignored by the authorities in the aftermath of the
flooding. Many were said to have been very upset at the loss of treasured possessions
and the disruption of their daily routines. This had led to behavioural problems and
increased anxiety levels during rainfall.
Parents wanted to be given advice on how to deal with children after flooding and the
sort of impacts they might face, and the availability of psychological or emotional
support for children in the aftermath of flooding. Creches and child-care facilities where
children could be left while parents were coping with the recovery period were also
suggested.
The flooding was seen to have had differential impacts upon men and women.
Women were seen to be particularly affected by the flooding, both physically and
psychologically. Single women were thought to have been taken advantage of by
insurance companies and builders.
Support for the elderly and disabled following the flooding was seen to be generally
inadequate.
Table 2. Physical health effects reported by focus-group participants
Todmorden West Auckland South Church
asthma coughs blood disorder
chest infections diarrhoea/ chest infections
upset stomachs
coughs/colds/flu kidney infection coughs
diarrhoea/vomiting/ laryngitis diarrhoea/
upset stomachs upset stomachs
headaches pleurisy high blood pressure
skin irritations kidney infections
spots panic attacks
swollen glands pleurisy
throat and ear infections viral infections
ple, one woman in Todmorden was in remission from cancer, one male suffered from
multiple sclerosis and another elderly man had osteo and rheumatoid arthritis. One
respondent’s wife in West Auckland had suffered a heart attack six weeks before the
floods; another woman who had been badly affected by previous flooding reported
suffering prolonged stress and ill-health since the event. Many of these people had
either seen their health conditions exacerbated after the June 2000 flooding, or had
been concerned that they would be.
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
Vulnerability to flooding: health and social dimensions 1517
Most of the focus-group participants had visited their local doctor following the
floods. Several spoke of difficulty in getting appointments due to the large numbers
of people in the same situation. Having to wait for a week or two did not appear
to be uncommon. The costs of repeat prescriptions had been high for one woman,
who said that she could ill-afford to pay for them. Many doctors had attributed their
patients’ health problems to the stress of the flooding and the living conditions with
which people had to cope. Several respondents spoke of doctors not being able to
state whether the flood had been the cause of their health problems. The idea of
being labelled as a ‘flood victim’ by her doctor had upset one woman who said that
it made her feel ‘like a leper’, while one South Church woman had not been to her
doctor for fear of losing her job owing to gaining a reputation for ill-health.
The flood-action group in South Church was concerned that the health effects of
the flooding would be ignored by the authorities because, according to the council,
there had been no cases of ‘disease’ reported by local GPs to the Health Authority.
The Chief Executive also stated that, although the Council were frequently in contact
with the communicable-diseases department and were getting reassurances that no
diseases had been reported, it is only serious diseases that are reported. Therefore,
people can be quite ill following floods but may not be reported as being so by their
doctors.
People’s ongoing concerns about their physical health were largely associated with
the perceived health risks from streets not being cleaned following the flooding (par-
ticularly where children were playing in these environments). Health risks from con-
taminated drinking water were also mentioned during the focus-group meetings.
Several people had sought advice on this from various sources and had largely been
reassured that there were no serious risks. However, associated with these concerns
was the fear of rats, and the danger to health from diseases that they may carry.
Several people were particularly worried about diseases such as hepatitis, typhoid
and Weil’s disease.
A further concern related to possible bacteria left in the building materials of
properties following the flooding. Some people were disturbed that they had not
been offered vaccinations against hepatitis, as the police had apparently been. There
was more concern when it was suggested by several people that the reason they had
not been offered these vaccinations was because mass vaccination was too costly. A
number of people stated that they would have been willing to cover any such costs
themselves.
One issue which was of concern to most of the people taking part in the focus-group
meetings was the lateness of advice on dealing with contaminated possessions. Flyers
were sent out by local authorities a day or two after the floods, advising people to
wash their hands if they were handling contaminated goods. Most people stated that
these notices arrived too late as they had immediately begun to clean their homes the
moment the flood waters had receded. Conflicting information and advice on health
issues given by different authorities was another area causing anger and concern.
A further concern to participants in West Auckland was whether the flooding had
contributed to increased mortality rates in the affected areas, albeit of those who
were already elderly or sick. There is evidence that death can be hastened by the
experience of flooding, rather than somehow being caused by it. Following the 1968
flooding in Bristol, Bennet (1970) reported that mortality rates for the 12 months
following the flood increased significantly in the homes of people who were flooded.
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
1518 S. M. Tapsell and others
(c) The mental health effects of flooding
The general conclusion from all six focus-group meetings was that everyone had
suffered from, or was continuing to suffer from, psychological health impacts from
the stress of the flooding. Stress is seen as arising from the difference between the
perceived demand the event places upon the individual and the resources the indi-
vidual can draw upon to adapt to that demand (Green 1988). The severity of the
impact represents the degree to which coping and support capacity are insufficient
to cope with the challenge and the costs of responding (figure 1). Moreover, the
stress associated with a particular disaster is not limited to the time that it occurs,
and both short-term and long-term stress reactions can be evident in disaster vic-
tims (Holen 1991; Boscarino 1997). Evidence from the Thames region study (Tapsell
2000) showed that the psychological effects of flooding can continue for months or
even years after the event, and are often more pronounced than the physical health
effects.
It is expected that people who experience a traumatic event will have a severe
reaction to it, particularly in the first few days after the event (e.g. Ticehurst et
al. 1996). For people who display symptoms beyond this period there are three
possible diagnoses: adjustment disorder, acute-stress disorder, and post-traumatic-
stress disorder. The three can be differentiated by the type of stressor, and the range
and duration of symptoms (Rick et al. 1998). Many people in the focus groups were
displaying symptoms of impaired mental health such as those related to adjustment
disorder (Rick et al . 1998). These symptoms included avoidance of talking or thinking
about the flooding, flashbacks, sleep disorders and depression.
Only a few people in the northeast England focus groups felt that they were now
‘over’ the flood, and these tended to be men rather than women. The chairperson of
the flood-action group stated that she ‘would not know where to begin’ to evaluate
the psychological impacts of the flooding on the community. It was suggested by par-
ticipants in South Church, as well as the chairperson of the flood-action group, that
substance abuse, such as reliance on alcohol and prescription drugs, had probably
risen since the flooding, or would increase in the future.
The nature of the trauma (e.g. flood event) may impinge upon the severity of
symptoms. Research suggests a complex relationship between the experience of a
traumatic event and subsequent mental disorders. Evidence from the focus groups
suggests that the greater the damage, losses and inconvenience from a flood event,
the greater the stresses suffered by flood victims are likely to be. Several people in
the groups stated that they could not cope with another flood and the associated
impacts it would have on them and their households. One couple felt that keeping
occupied with the recovery process had helped them cope with the event.
There was often no escape from the stresses of recovery. One female participant
in Todmorden spoke of the relief she had felt when able to take a brief holiday.
On the other hand, there were some positive outcomes from the flooding for a few
participants in the focus groups. One older woman felt that the experience had
helped her assess some long-term childhood anxieties, and two other women said
that they had become good friends since the flooding and that many people now
knew neighbours with whom they had not previously been acquainted. Several people
spoke of being able to laugh at their experiences and that a sense of humour helped
them to cope with what had happened.
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
Vulnerability to flooding: health and social dimensions 1519
Following a flood, subsequent rainfall can act as a trigger for repeated stress and
trauma among those who were flooded, being a factor directly related to the ini-
tial flood event. Most focus-group participants in all the three locations we studied
expressed concern and anxiety during heavy rainfall since the flooding, matching the
findings from our previous studies (e.g. Tapsell et al . 1999). Increased awareness of
the hazard meant that some people had begun to differentiate between the various
intensities of rainfall and had started to refer to these different types. This reflects
the situation in the Thames region study where people had began to refer to ‘Easter
rain’ to differentiate it from ‘normal’ rainfall.
(d) Changes in behaviour since the flooding
Many people had changed aspects of their behaviour since the flooding. This
ranged from moving possessions to higher parts of the house to refurbishing the
property differently in case of future flooding. These same behavioural patterns were
also noted in the Thames region study.
Disasters and their aftermath can have significant impacts upon people’s relation-
ships, both within the household and with those outside. A few people in the focus
groups spoke of positive changes in relationships with their partners as a consequence
of the flooding, but some mentioned relationship problems in their households and
in the community due to the stress of the flooding and the recovery process.
When people were asked what had been the most devastating aspect of the flooding
at a personal level, responses varied from citing financial impacts upon households,
to ‘lost’ time, disruption of family life, losing everything that people had worked
for, and the loss of sentimental possessions. Some people referred to the feeling of
powerlessness and helplessness and not being able to do anything to stop the flood
and save their homes. The chairperson of the flood-action group found it difficult to
say which of the many serious impacts of the flooding on people had been the worst:
for herself and others in the action group, dealing with the flood had literally ‘taken
over their lives’.
3. A proposed So cial Floo d Vulnerability Index (SFVI)
Given an increasing appreciation of the full impacts of floods, from research such as
reported above, we have sought to develop an index which measures the impact that
floods could have upon the communities potentially affected. We anticipate that
this index would be used in addition to knowledge of the potential flood damage
impacts and losses that are more readily understood and modelled within benefit–
cost analyses within investment appraisals. We outline here, therefore, the definition
and source of the variables used in the calculation of the Flood Hazard Research
Centre (FHRC) SFVI, which can be used to predict those areas and populations
that are likely to be most severely affected in terms of health and other ‘intangible’
flood impacts.
Our choice of data was constrained by the need, first, to use secondary source data
that are available for the whole of England and Wales, rather than rely on expensive
and time-consuming interview or focus-group approaches. We also needed, secondly,
to use data that are available for small geographical areas, because flood plains are
often narrow and short. We used 1991 census data from the Manchester Information
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
1520 S. M. Tapsell and others
and Associated Services (MIMAS), because these data fit the above criteria, being
available for England and Wales at the level of the enumeration district (ED).
(a) The SFVI and its constituents
The SFVI is a composite additive index based on three social characteristics and
four financial-deprivation indicators. The rationale for the selection of these vari-
ables is given in table 3, and figure 2 maps these and other variables against the six
case studies that we have undertaken over several years to quantify the ‘intangible’
impacts of flood events. The linkages shown in figure 2 are based on our best judg-
ment, supported by results from the hundreds of interviews in those studies, of the
relative effect of the different variables in determining household vulnerability in the
types of floods experienced in those localities. It shows that the age and financial sta-
tus of the affected populations are the most commonly important variables, followed
by the prior health status of the population. We have chosen to use the incidence
of lone parents as a measure of family structure, because our research points to this
causing extremes of vulnerability, although the presence or absence of young children
(to which the incidence of lone parents is very closely correlated) had, perhaps, an
equal case for inclusion.
To identify the financially deprived, the Townsend Index (Townsend et al. 1988)
was used because, unlike other deprivation indices, it focuses on deprivation out-
comes (such as unemployment), rather than targeting predefined social groups. This
enabled us to identify our own social classification. This is important because finan-
cial deprivation is only one of several factors that contribute towards vulnerability
to flood impacts and it is our intention to target only those social groups which
previous research has shown to be particularly badly affected.
The Townsend indicators are as follows.
(1) Unemployment: unemployed residents aged 16 and over (S090019+S090043)
as a percentage of all economically active residents aged over 16 (S090013 +
S090037).†
(2) Overcrowding: households with more than one person per room (S230003 +
S230004) as a percentage of all households (S230001).
(3) Non-car ownership: households with no car (S210003) as a percentage of all
households (S210002).
(4) Non-home ownership: households not owning their own home ((S200001 +
S200009) − (S200002 + S200003)) as a percentage of all households (S200001 +
S200009).
In order to prevent any undue bias in the SFVI towards financial deprivation, the
four Townsend indicators were summed and multiplied by 0.25 before being added to
the other variables. Those other variables and the social groups that they highlight
are as follows.
(1) The long-term sick: residents suffering from limiting long-term illness
(S120001) as a percentage of all residents (S010064).
† These codes refer to the tables provided by MIMAS. Hence S090019 refers to census Small Area
Statistics, table 09
, variable 0019.
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
Vulnerability to flooding: health and social dimensions 1521
indicators not
used in the SFVI
studies
and dates
indicators used
in the SFVI
gender young children ethnicity
Town &
Aberconwy
(1999)
6
Euroflood
(miscellaneous
UK locations)
(1994)
5
Banbury &
Kidlington
(1999)
4
North East
of England
(2001)
3
Strathclyde,
Perth &
North Wales
(1995)
2
Uphill
(1985)
1
lone parents aged over 75 financially deprived long-term sick
weak linkage
moderate linkage
strong linkage
key
Figure 2. The relationship between results from a range of case studies and the choice of vari-
ables for the SFVI (based on qualitative social science methods in focus-group research, without
control groups). References labelled in the figure are as follows: 1, Green et al. (1985b); 2, Ford-
ham & Ketteridge (1995) and Hill & O’Brien (1999); 3, Tapsell & Tunstall (2001); 4, Tapsell et
al. (1999); 5, Ketteridge & Green (1994); 6, Hill & O’Brien (1999).
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
1522 S. M. Tapsell and others
Table 3. Our rationales for the selection of variables for the SFVI
(based on qualitative focus-group research)
variables rationale
elderly
(aged 75+)
The age of 75 was chosen because epidemiological research has shown
that after this age there is a sharp increase in the incidence and
severity of arthritis (and other conditions) and this illness is sensitive
to the damp, cold environmental conditions that would follow a flood
event.
lone parents Previous FHRC research has shown that lone parents (of either sex)
are badly affected by floods because they tend to have less income
and must cope singlehandedly with both children and the impact of
the flood, with all the stress and trauma that this can bring (Green et
al. 1987).
preexisting
health
problems
Research by FHRC has shown that post-flood morbidity (and
mortality) is significantly higher when the flood victims suffer from
preexisting health problems (Green et al. 1994).
financial
deprivation
The financially deprived are less likely to have home-contents
insurance and would therefore have more difficulty in replacing
households items damaged by a flood event (and it would take longer).
Table 4. Transformation methods used in the compilation of the SFVI
indicator transformation method
lone parents log natural (x +1)
aged 75+ log natural (x +1)
long-term sick square root
non-homeowners square root
unemployed log natural (x +1)
non-car owners square root
overcrowding log natural (x +1)
(2) Single parents: lone parents (S400001) as a proportion of all residents
(S010064).
(3) The elderly: residents aged 75 and over (S020127 + S020134 + S020141 +
S020148) as a percentage of all residents (S010064).
(b) Data processing
The crude percentages were transformed by the method that produced the mini-
mum skewness and kurtosis within their distributions. The selected transformation
methods are shown in table 4. Following transformation, the data were standard-
ized as Z-scores and then summed (as stated above, the Townsend indicators were
summed separately and then multiplied by 0.25 prior to their inclusion in the index).
The completed SFVI has a minimum value of −10.4, a maximum value of 9.07, a
mean of 0.06 and a standard deviation of 2.56. Given the relatively exploratory nature
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
Vulnerability to flooding: health and social dimensions 1523
Environment Agency indicative floodplain extents
River Irwell, Manchester
River Thames, Maidenhead
River centre lines
social vulnerability
very low
low
average
high
very high
(a)(b)
100 m
100 m
Figure 3. The results for the SFVI for (a) Salford, Manchester, and (b) Maidenhead.
of this work, the SFVI was then categorized into a limited number of bands (five),
where category (1) represents low vulnerability, category (3) average vulnerability
and category (5) high vulnerability, etc. Also, out of 113 465 EDs in England and
Wales there are 4982 missing values. This is due to either null values in the indicators
or, in some cases, data for the ED are ‘suppressed’. Data for a ward or ED are
suppressed when its population is considered to be so low that individual households
or persons could be identified from the census data that are released.
Figure 3 gives the results for two areas of the UK where flood-alleviation schemes
have been implemented in the recent past, showing for Maidenhead that the flood
plains are populated by relatively affluent communities with slightly lower average
SFVI values than in the surrounding areas. The Manchester communities, by this
measure, are much more vulnerable to flooding than their Maidenhead equivalents,
with a large vulnerable community in the southwest of the flood-plain area (in the
centre of figure 3a). Further research is needed to gauge how accurately actual vulner-
ability to flooding is predicted by these SFVI scores. That research should perhaps
both attempt to ‘hindcast’ the existing focus-group research results from the index
values and assess the actual vulnerability of populations affected by future floods in
comparison with the predicted scores.
4. Conclusions
The conclusions from this research are simply that the impacts of flooding are more
extensive and complex than have hitherto been appreciated, and that assessments
of the effect of flood-defence measures on reducing these impacts are flawed if only
monetary losses are used within the necessary project-appraisal methods. If flooding
is to become more frequent or more severe under regimes of altered climatic condi-
tions, governments will need to allocate even more resources to this area of public
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
1524 S. M. Tapsell and others
expenditure. There is no failure to appreciate this now in government circles, fol-
lowing major flooding in 1998 and 2000; the test will come when these floods are
forgotten by those who were not directly affected, and decision-makers move on to
other investment priorities within the public domain.
References
Bennet, G. 1970 Bristol floods 1968: controlled survey of effects on health of local community
disaster. Br. Med. J. 3, 454–458.
Boscarino, J. A. 1997 Diseases among men 20 years after exposure to severe stress: implications
for clinical research and medical care. Psychosom. Med. 59, 605–614.
Environment Agency 2000 June 2000 floods: report to the Minister (northeast region). Environ-
ment Agency Northeast Region, Leeds.
Fordham, M. & Ketteridge, A. 1995 Flood disasters—dividing the community. Enfield: Flood
Hazard Research Centre, Middlesex University.
Green, C. H. 1988 The relationships between the magnitude of flooding, stress and health. Flood
Hazard Research Centre, Middlesex University, Enfield.
Green, C. H., Emery, P. J., Penning-Rowsell, E. C. & Parker, D. J. 1985a The health effects of
flooding: a survey at Uphill, Avon. Flood Hazard Research Centre, Middlesex Polytechnic,
Enfield.
Green, C., Emery, P., Penning-Rowsell, E. & Parker, D. 1985b Evaluating the ‘intangible’ effects
of flooding on households: a survey at Uphill, Avon. Flood Hazard Research Centre, Middlesex
Polytechnic, Enfield.
Green, C. H., Penning-Rowsell, E. C. & Parker, D. J. 1987 Estimating the risk from flooding and
evaluating worry. In Uncertainty in risk assessment, risk management and decision making
(ed. V. T. Covello, L. B. Lave, A. Moghissi & V. R. R. Uppuluri). New York: Plenum.
Green, C. H., Van der Veen, A., Wierstra, E. & Penning-Rowsell, E. 1994 Vulnerability refined:
analysing full flood impacts. In Floods across europe: flood forecasting, assessment, modelling
and management (ed. E. C. Penning-Rowsell & M. Fordham). London: Middlesex University
Press.
Hill, J. & O’Brien, P. 1999 Disaster in the community: emergency planning for sustainable
solutions to long-term problems. Worker and resident perspectives of the North Wales floods
1990 and 1993. Disaster Recovery and Research Team Ltd, Caernarfon, Gwynedd.
Holen, A. 1991 A longitudinal study of the occurrence and persistence of posttraumatic health
problems in disaster survivors. Stress Med. 7, 11–17.
Ketteridge, A. & Green, C. (eds) 1994 The technical annex for the full flood impacts module.
Flood Hazard Research Centre, Middlesex University, Enfield.
Miles, M. B. & Huberman, A. M. 1994 Qualitative data analysis, 2nd edn. London: Sage Publi-
cations.
Ohl, C. A. & Tapsell, S. M. 2000 Flooding and human health: the dangers posed are not always
obvious. Br. Med. J. 321, 1167–1168.
Parker, D. J., Green, C. H. & Penning-Rowsell, E. C. 1983 Swalecliffe coast protection
proposals—evaluation of potential benefit. Flood Hazard Research Centre, Middlesex Poly-
technic, Enfield.
Penning-Rowsell, E. C. & Green, C. H. 2000 New insights into the appraisal of flood alleviation
benefits. I. Flood damage and flood loss information. J. Inst. Water Environ. Manag. 14,
347–353.
Rick, J., Perryman, S., Young, K., Guppy, A. & Hillage, J. 1998 Workplace trauma and its
management: review of the literature. Report for the Health and Safety Executive. Institute
for Employment Studies, University of Sussex, Brighton.
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
Vulnerability to flooding: health and social dimensions 1525
Tapsell, S. M. 2000 Follow-up study of the health effects of the 1998 Easter flooding in Banbury
and Kidlington. Report to the Environment Agency, Thames region. Flood Hazard Research
Centre, Middlesex University, Enfield.
Tapsell, S. M. & Tunstall, S. M. 2001 The health and social effects of the June 2000 flooding
in the northeast region. Report to the Environment Agency. Flood Hazard Research Centre,
Middlesex University, Enfield.
Tapsell, S. M., Tunstall, S. M., Penning-Rowsell, E. C. & Handmer, J. W. 1999 The health
effects of the 1998 Easter flooding in Banbury and Kidlington. Report to the Environment
Agency, Thames region. Flood Hazard Research Centre, Middlesex University, Enfield.
Ticehurst, S., Webster, R. A., Carr, V. J. & Lewin, T. J. 1996 The psychosocial impact of an
earthquake on the elderly. Int. J. Geriat. Psychiatry 11, 943–951.
Townsend, P., Phillimore, P. & Beattie, A. 1988 Health and deprivation: inequality and the
North. London: Croom Helm.
World Health Organization 1948 Constitution of the World Health Organization. Geneva: World
Health Organization.
Phil. Trans. R. Soc. Lond. A (2002)
on May 24, 2011rsta.royalsocietypublishing.orgDownloaded from
