The Scottish Government

« Previous | Contents | Next »

Listen

CHAPTER FOUR HOUSEHOLD SURVEY

4.1 The questionnaire survey was delivered to 2,321 households in

Brechin, Edinburgh, Elgin, Forres, Glasgow-Shettleston, Hawick and Perth (Table 4.1 and Figure 4.1). Households were randomly selected within historic flood envelopes provided by the relevant local authority. This meant that all surveyed households were at risk, but due to local topographic effects or changed occupancy since the flood, not all households had been flooded. The locations of the major urban communities sampled in the household survey can be inspected in Ordnance Survey 1:50,000 map extracts of Brechin, Edinburgh, Elgin, Forres, Glasgow-Shettleston, Hawick and Perth in Appendix A. The flood outlines on these maps are those from the Indicative River & Coastal Flood Map (Scotland) published by SEPA in November 2006. This was not available at the time of the household survey (March - June 2006) but it does provides a general indication of the sampled communities within the flood-risk areas.

4.2 Responses from 1,223 households were obtained giving an overall response rate of 52.7%. The bulk of the sampling was undertaken by a door-to-door survey, questionnaires being delivered one day and generally collected the next. The scattered nature of the target population for rural and coastal areas (236 households in Inverness, Mentrie, Eyemouth, Dunoon, Copach, Highland region, Orkney and Shetland, Figure 4.1) made a door-to-door survey too costly and a postal survey was adopted instead. The response rate of 29.2%, whilst lower than for the door-to-door survey, is still acceptable for surveys of this kind.

Table 4.1 Response rates by survey location

Socio-demographic and residential characteristics of the sample

4.3 The sample yielded 633 flooded properties and 590 non-flooded properties ( Appendix B; Table 2) generating a close to even split between respondents who had been flooded (51.8%) and those who lived in a flood risk zone (48.2%).

Age and occupation

4.4 The age profile within households ( Appendix B; Table 3) broadly reflects national patterns; 22.2% of households include at least one adult more than 70 years old. The occupational groups of the highest earner are also broadly in keeping with the national profile (Table 4.2). For both 'age' and 'occupation' the sub-samples are sufficiently large to enable differential responses to flooding and flood risk to be examined at a later stage.

Figure 4.1 Sample sites used in the household survey

Table 4.2 Occupational group of highest earner in household

Housing tenure and property type

4.5 In terms of housing tenure (Table 4.3), the majority of the properties are privately owned (75.6%) with rented accommodation provided by local authorities (15.8%), housing associations (4.9%) and private landlords (3.1%). This profile also broadly conforms to the national pattern.

Table 4.3 Housing tenure

4.6 The two dominant types of housing stock ( Appendix B; Table 6) comprise terraces and tenements (63.1%), followed by detached and semi-detached properties (35.3%). Groundfloor only bungalows and flats made up 43.1% of the properties sampled and houses with both ground and upper floors 49.2% ( Appendix B; Table 7). The remaining 7.7% of properties comprised upper floor flats which had been flooded, below to the agreed ceiling of 10% (para. 3.14).

Location

4.7 The majority of properties (49.2%) were located in small urban areas (Brechin, Elgin, Forres and Hawick) with a further 45.4% drawn from large urban areas (Glasgow, Edinburgh and Perth). The balance of 5.6% was drawn from scattered inland and coastal rural locations ( Appendix B; Tables 1 and 8).

Characteristics of the floods

4.8 Inundation by rivers overtopping their banks (fluvial flooding) was the single most important source of flooding (85.4%), followed by urban flooding (pluvial) caused by surcharging sewers combined with overland flow (9.9%). Coastal flooding was only recorded in 57 properties (4.7%), but this does represent a significant proportion of properties recently subject to coastal floods. The categorisation of flooding type was made by project staff, reflecting information provided by SEPA staff and local authorities in their Biennial Reports for specific events in the sampled locations (Table 4.4).

Table 4.4 Types of flood

It is recognised that a clear distinction between fluvial and pluvial floods is not always possible. In this study pluvial flooding relates primarily to the Shettleston event in the east end of Glasgow in 2002. The occurrence, severity and spatial extent of different types of flood over the last decade has constrained our ability to include more properties subject to pluvial and coastal flooding.

4.9 In properties which had been flooded, 60% reported water over one foot deep (30 cm) and only 6.3% less than one inch deep (2.5cm), implying that most floods generate significant depths at ground floor level ( Appendix B; Table 10). In 48.5% of properties sampled, the ground floor was inundated and in 16.4% of properties inundation only extended to the gardens and outhouses or garages ( Appendix B; Table 2).

4.10 In 81.4% of events the flood waters contained mud, silt or gravel ( Appendix B; Table 11) and in 59.4 % sewage ( Appendix B; Table 12). All types of floods (fluvial, pluvial and coastal) contained significant amounts of sediment but, as one might expect, sewage was more prevalent in pluvial (73.3%) than in fluvial floods (57%).

4.11 Of those who had been flooded, 74% had experienced flooding in their homes or communal stair and 26% only in their gardens and/or outhouses ( Appendix B; Table 2). Of the 629 respondents who had been flooded 71.7% had experienced one flood, 22.7% two floods and 5.6% three or more floods ( Appendix B; Table 13).

The experience of being flooded

4.12 Of those who were flooded, 88% of respondents were at home when the flood occurred. In addition to noting rising water, in 42% of properties a flood warning was received mainly via neighbours (31.8%) or an official (27.5%). For 273 respondents (43% of the sample) Floodline accounted for only 8.1% of warnings. ( Appendix B; Table 14). Overall, 33.4% of households received a warning at least three hours before their property became inundated ( Appendix B; Table 15).

4.13 Priority actions prior to the flood ( Appendix B; Table 16) comprised removing possessions from the ground floor (39.3%), deploying sandbags or flood guards (37.4%), followed by moving the car (33.9%) and vacating the house (21.3%). Of the 86.8% of respondents who undertook mitigation measures in response to a warning (Table 4.5), 49.2% removed possessions from the ground floor, 45.9% deployed sandbags or flood guards, 39.7% moved a vehicle to higher ground and 28.9% evacuated members of the household. Only 62.5% of those not warned undertook the same mitigation measures, but at much lower levels of activity. In general, the flood warning generated a higher level of activity in each of these priority actions.

4.14 When priority actions are cross-tabulated against the length of the warning lead time ( Appendix B; Table 18), actions within the house were scaled down after three hours, possibly reflecting physical exhaustion and/or evacuation of the property.

Table 4.5 Actions taken prior to the flood, by whether received flood warning (per cent of respondents)

4.15 By far the largest source of assistance ( Appendix B; Table 19) was from neighbours (55.8%) and family (53.3%), followed by friends outside the locality (27.9%), the local authority (21%) and the Fire Service (19.7%). Of those respondents who had been home when flooded, 58% assisted their neighbours. A further 44% of those at home, but not flooded, also provided assistance.

4.16 The majority of those forced to leave their home (Appendix B; Table 20) stayed with friends/relatives (44.2%) and/or rented alternative accommodation (52.4%). Very few (4.8%) reported using an Evacuation Centre, but this may reflect in part the subsequent use of alternative longer-term accommodation. Many moved from one type of accommodation to another during the period in which they were displaced from their homes.

Impacts of flooding

4.17 Once flooded, 78% of those affected ( Appendix B; Table 21) vacated their homes and, of these, 45.4% were out their homes for six months or more, the remaining 54.6% returning within six months.

4.18 Direct economic losses sustained by individual households ( Appendix B; Tables 22 and 23) comprised damage to buildings and contents (mean values of £31,980 and £13,552 respectively, though in many cases these losses would be offset by insurance) and unpaid leave averaging 10.4 days. An average of 6.3 days annual leave was also taken in the wake of flooding. The average length of compassionate leave taken from work (with the cost being met by employers) was 9.8 days.

4.19 In order to gauge the impacts of flooding on households, respondents were invited to score 20 potential impacts on a scale: 'not applicable', 'no impact', 'mild', 'serious' and 'extreme impact' ². The overall results in terms of respondents reporting a serious or extreme impact are shown in rank order in Figure 4.2 and listed in Table 4.6.

Figure 4.2 Categories of flood impacts ranked according to severity

4.20 Each category can also be assessed in terms of its average score (0-3). Time and effort to return to normal (2.21) and discomfort/inconvenience (2.20) are clearly identified as the most severe impacts, closely followed by worry about future flooding (2.14), stress of the flood itself (2.03) and having to leave home (1.91). Impacts directly related to being out of one's home are then registered including dealing with insurers (1.67) and builders (1.62), living in temporary accommodation (1.62) and being stranded in/out of home (1.62). Financial loss (1.53) together with loss of house value (1.36) and damage to car (0.76) occupy the second half of the bar chart alongside losses of irreplaceable/sentimental items (1.53), deterioration in physical (0.82) and mental health (0.94), strains between family members (1.02) and loss of holiday entitlement (0.65). Other impacts that score relatively modestly include disruption to electricity (1.33), loss of community spirit (0.97) and loss or distress to pets (0.71); although the last of these may be depressed by responses from households without pets.

4.21 In an attempt to probe deeper in terms of flood impacts, each of the 20 categories has been assigned to one of three groups: tangible (relating to material losses) and intangible (relating to non-material and/or emotional losses). The latter group is further sub-divided into immediate impacts and lasting impacts (Table 4.6). Tangible impacts are losses that can readily be expressed in material terms (for example, loss in house value, damage to buildings contents and vehicles resulting in a measurable financial loss). By contrast, intangible impacts typically involve affective responses. They are less readily convertible into costs although monetised values can be attributed to them by economists (for example, see JBA (2005) and RPA (2005)). Intangible impacts can be further divided into those whose impact is felt during and immediately after the flood (for example, being forced to evacuate, the stress of being flooded, dealing with insurers and builders) and impacts that are delayed but lasting and only surface several months later (for example, worry about future flooding, loss of irreplaceable family pictures and memorabilia, deterioration in physical and mental health). It is recognised that this typology is not absolute and there are gradations between impacts that are tangible or intangible and immediate or lasting. Nevertheless, we have found this typology helpful in categorising the range of impacts experienced by households that have been flooded.

4.22 Average scores across all households for each of the new classes of impacts have been derived in Table 4.6 for the whole population sampled ('no response' included in the calculation) and for only those who reported an impact ('no response' excluded from the calculation). The former is valid when generalising impacts across the whole population which experienced flooding: the latter is valid if only those who reported an impact (zero, mild, serious or extreme) are included. For example, across the whole sampled population 'loss or distress to pets' scores very low (0.71) whereas amongst pet-owners this impact increases nearly two-fold (1.45). For the sampled population the scores are tangible impacts 1.12, intangible immediate impacts 1.81 and intangible lasting impacts 1.29. Two points immediately stand out: intangible impacts register markedly higher values than tangible impacts and the immediate impacts are, on average, strikingly higher than the lasting impacts. These findings undergo some adjustment when only households which reported an impact are included. The rank order across categories remains the same, but the scores change differentially. Tangible impacts are now 1.52 (up 0.4), intangible immediate impacts 2.07 (up 0.26) and intangible lasting impacts 1.57 (up 0.28). For households that register a flood impact, average tangible impacts approach intangible lasting impacts, but intangible immediate impacts are markedly higher than either. This finding is also apparent in Figure 4.2 with individual serious or extreme tangible impacts recorded by 19-50% of households whereas for intangible immediate impacts the range is 51-82% of households.

Table 4.6 Flood impacts

* 'N/A' or 'No impact'=0; 'Mild impact'=1; 'Serious impact'=2; 'Extreme impact'=3
** 'No impact'=0; 'Mild impact'=1; 'Serious impact'=2; 'Extreme impact'=3

4.23 The composition of intangible immediate impacts across the sampled population (column 2 in Table 4.6) all relate to the stress of the flood itself (2.03), the anxiety caused by being out of one's home (1.91), the discomfort of living in temporary accommodation (2.20) and time and effort spent dealing with insurers (1.67) and builders (1.62) in order to return home. Although labelled 'immediate' this group of impacts often extends over many months (45.4% of households were out of their home for six months or more, Appendix B; Table 21). Given the primacy of the goal of returning home, it comes as no surprise that this group of impacts scores so high. However, once the household has returned to its home, many of these impacts will diminish in significance - hence the use of the term 'immediate'. When only households which registered an impact are included in the sample, the absolute values of impacts increase but, apart from having to leave home (markedly up) and having to deal with insurers (markedly down), their rank order is unchanged (Table 4.6, column 4).

4.24 By contrast, the individual elements that collectively comprise intangible lasting impacts register effects that have a sustained and longer term impact within households across the sampled population. The high score attributed to the time and effort required to return to normal (2.20) points to the flood as a highly disruptive element within the life of the household. Part of that return to 'normal life' is hindered by worry of future flooding (2.14), the loss of sentimental/irreplaceable items (1.53) and strains within the family (1.02). Concerns over physical and mental health do not score so highly (0.82 and 0.94 respectively), but this may possibly reflect the stigma attached to scoring mild or serious impacts within such categories. When viewed alongside strains within the family and worry about future flooding, the mental health impacts of flooding could be more severe than is implied by the self coding of the questionnaire (see para 5.18). When only households which reported an impact are included the absolute value always increases, but the rank order is little changed with only loss or distress to pets and deterioration in mental registering higher ranks than before (Table 4.6, column 4).

4.25 Tangible impacts for the sampled population recorded the lowest group average score (1.12). None of the individual impacts scored especially high with financial loss (1.53) and loss of house value (1.36) and loss of electricity (1.33) registering the highest values. All values increase when only households reporting these impacts are included, but the rank order remains unchanged (Table 4.6). It is striking that this group of impacts, which reports material losses most easily measured in economic terms, scores the lowest aggregate score. This may reflect the relatively high take up of insurance, with 94.7% having contents insurance ( Appendix B; Table 43).

4.26 We now explore flood impacts on households, sub-dividing the responses according to their experience of the flood (warning received, insurance status, age profile, occupational status, household income, number of times flooded, type of flood, depth of flood, location and housing tenure).

Warning

4.27 Receiving a flood warning is associated with a lower average score in terms of time and effort to return to normal (2.18 compared with 2.24 for households with no warning), reducing worry about future flooding (2.10 compared with 2.17) and the loss of irreplaceable items (1.48 compared with 1.57) and lower overall financial losses (1.48 compared with 1.57). All of these differences are modest, and it is interesting that the largest differences in this group (loss of irreplaceable items and overall financial losses) are those for which residents can take action to reduce flood impact. Overall average intangible-immediate impacts are slightly reduced (1.79 with a warning and 1.83 without a warning) but are unchanged for overall average intangible lasting impacts ( Appendix B; Table 25). Tangible impacts also report a slight reduction when a warning is received (1.11 versus 1.13).

Insurance

4.28 The effect of having contents insurance on reported flood impacts is complex. The questionnaire was designed primarily to check on contents insurance, it being recognised that many living in rented accommodation would be unsure as to their buildings cover (in Appendix B; Table 22 only 184 reported the actual value of buildings losses). However in practice, properties being bought via a mortgage, or properties rented from a local authority, will be covered by buildings insurance - as will the majority of privately owned non-mortgaged properties. Thus the term "contents insurance" can broadly be interpreted as including both buildings and contents. As expected, dealing with insurers (1.95 with insurance, 1.88 without) and builders (1.73 with insurance and 1.00 without) register higher impacts when contents insurance is held, as does living in temporary accommodation (1.68 with insurance, 1.50 without). Discomfort/ inconvenience (2.27 with insurance and 1.97 without) and time and effort to return to normal (2.28 with insurance and 2.05 without) also register more severe impacts despite having contents insurance. This may reflect higher levels of affluence in households with contents insurance with higher claims generating additional and extended post flood adjustment. This is partly confirmed by higher tangible impacts and higher intangible-immediate impacts being reported by those with contents insurance ( Appendix B; Table 26).

Elderly household members

4.29 The presence of at least one member of the householder more than 70 years old has little effect on average individual impacts. In only two categories does the presence of a septuagenarian strongly increase the average score: having to leave home (1.98 compared with 1.89) and being stranded in/out of home (1.73 compared with 1.58). However, it is striking that all three groups of impacts score lower averages when a septuagenarian is a member of the household ( Appendix B; Table 27). A possible explanation is that the handling many post flood impacts are moderated by family and/or friends.

Income level

4.30 Some of the largest differences in response emerge when the sample is sub-divided into households with annual incomes above and below £20,000. For those below this income threshold, the stress of the flood (2.02 compared with 1.83), having to leave home (1.95 compared with 1.70) and the fear of being stranded in/out of one's home (1.61 compared to 1.26) generate higher levels of anxiety. To this should be added the higher impact of losing irreplaceable items (1.56 compared to 1.23) plus increased strains within the family (1.02 compared with 0.88), more severe mental (1.07 compared with 0.64) and physical health effects (0.92 compared with 0.47) and a view that community spirit has been weakened since the flood (0.86 compared with 0.74). This profile of responses points to vulnerable households where modest incomes may inhibit post flood recovery. It is also striking that in the aggregate scores ( Appendix B; Table 28) tangible impacts are not income sensitive, but both immediate and lasting intangible impacts are more severe for lower income households.

Occupational status

4.31 When households are grouped according to the occupational status of the highest earner ( Appendix B; Table 29) there is a general pattern of average tangible impacts being more severe in skilled or semi-skilled households (1.19) when compared with professional or managerial households (1.10). A similar pattern obtains for intangible lasting impacts (1.37 compared with 1.16). For immediate intangible impacts, occupational status is a poor discriminator.

Frequency of flooding

4.32 We can also explore whether repetition intensifies or lessens individual flood impacts, by comparing responses from individuals flooded once, twice or three or more times. In fifteen out of seventeen instances the impact of being flooded a second time was more severe with individual scores increasing by as much as 0.54 (worry about future flooding and loss in house value). The impacts following a third flood fall back somewhat, possibly reflecting a fatalistic reaction, but these results are based on much smaller samples with associated reductions in confidence. The contrast in average scores is marked and noteworthy ( Appendix B; Table 30) with tangible impacts up from 1.05 to 1.34, intangible immediate impacts up from 1.74 to 2.06 and intangible lasting impacts up from 1.20 to 1.55 on being flooded a second time. Repetition would appear to intensify the pain especially in terms of lasting impacts.

Flood type

4.33 Differential impacts in terms of flood type and depth of flood waters can also be examined, although the small number of returns from those affected by coastal floods precludes their separate analysis. Pluvial flooding generated strikingly more severe impacts in terms of financial loss (1.97 compared with 1.47 for fluvial floods), more significant losses of irreplaceable items (1.80 compared with 1.50), the stress generated by the flood (2.29 compared with 2.01), worry about future flooding (2.37 compared with 2.12), strikingly increased strains within the family (1.33 compared to 0.98), and greater deterioration in mental health (1.32 compared with 0.90) and physical health (1.14 compared with 0.79). In general, pluvial floods are characterised by less warning and higher levels of sewage contamination than fluvial floods - both of which are likely to generate higher tangible impacts and strikingly higher lasting intangible impacts ( Appendix B; Table 31). However, care must be taken in generalising from these findings as the sample size for responses for pluvial floods was relatively small (56) and mainly related to the Glasgow-Shettleston flood in 2002. The socio-demographic profile of the flooded population in Shettleston may be responsible for some of the reported differences in response.

Flood depth

4.34 As expected, the impact of a flood is strongly correlated with the depth of the flood water (Table 4.7). Whereas average group scores for a depth of 1 inch (2.5 cm) for tangible and intangible immediate impacts are 0.88 and 1.38 respectively, these increase to 1.31 and 2.05 when the flood water depth exceeds 12 inches (30 cm) - the level experienced by 60% of respondents. Within individual impacts, loss of irreplaceable items (up from 0.61 to 1.93), having to leave home (up from 1.35 to 2.20), the stress of the flood itself (up from 1.68 to 2.24) and living in temporary accommodation (up from 0.97 to 1.89) register substantially more severe impacts once flood waters exceed 12 inches.

Table 4.7 Flood impacts, by depth of flood (mean score*)

* 'N/A' or 'No impact'=0; 'Mild impact'=1; 'Serious impact'=2; 'Extreme impact'=3

Housing tenure

4.35 Housing tenure also has a significant effect on flood impacts (Table 4.8). Council house tenants typically register lower tangible impacts, but much higher lasting intangible impacts than owner occupiers. Especially severe impacts for council tenants include the stress of the flood itself, having to leave home, living in temporary accommodation, the fear of being stranded in/out of one's home, overall financial losses (and in particular the loss of irreplaceable items) and deterioration in both physical and mental health. For owner occupiers flood impacts were typically less severe except for dealing with insurers and builders, and concerns over loss of house value. The overall impact of a flood on council tenants appears to be more onerous and longer lasting when compared with owner occupiers, raising questions as to the relative vulnerability of both groups.

Location

4.36 Location also moderates the severity of flood impacts, in part reflecting the character of the most severe flood experienced by each household ( Appendix B; Table 34). In terms of tangible impacts, the floods in Elgin (1.32), Perth (1.27) and Glasgow-Shettleston (1.25) were the most severe and those in Hawick (0.81) and Brechin (0.74) the least severe. The same rank order obtains for immediate intangible impacts (Elgin 2.14, Perth 1.98 and Glasgow-Shettleston 1.92). But Glasgow-Shettleston (1.60) records the second highest value for lasting intangible impacts, very similar to that for Elgin (1.67). Overall, the experience of flooding in Elgin generates the highest average scores, closely followed by Glasgow-Shettleston.

Table 4.8 Flood impacts, by housing tenure (mean score*)

*'N/A' or 'No impact'=0; 'Mild impact'=1; 'Serious impact'=2; 'Extreme impact'=3

4.37 The overall impact of flooding on community cohesion is unclear with 17.1% of respondents reporting an improved sense of community and 15.6% reporting a deterioration ( Appendix B; Table 35). Further analysis may be needed to clarify any perceived lessening in community cohesion, especially if non-flood related drivers are involved.

4.38 When questioned on reasons for moving or considering moving (Table 36), 44.1% of respondents now saw their property as a poor investment, 83.2% feared another flood and many attached negative qualities to their house ( i.e. the house will never be the same and the house brought back negative feelings about the flood).

Living with floods

Risk perception

4.39 Given that all respondents live within or close to an area that has recently been flooded, it is striking that 63.6% of those who have been flooded view future flooding as "very likely" or "likely" whereas this reduces to 41.2% for those who have not been flooded (Table 4.9). The experience of being flooded significantly increases the perceived likelihood of future flooding. The perception of flood risk also significantly varies with location ( Appendix B; Table 38) with 88% of the sampled residents in Elgin viewing future flooding as "very likely" or "likely" (perhaps partly as a function of repeat flooding), this reducing to 44.4% in Brechin and 22.9% in Perth (now protected by a flood alleviation scheme).

Table 4.9 Perception of flood risk, by whether flooded (per cent of respondents)

Chi-squared = 86.532; df=4; p<0.001

4.40 The experience of being flooded does not result in reported improved knowledge of flood risk ( Appendix B; Table 39) with 67.8% of those flooded claiming to be either "not very well" or "not at all informed", a value very similar to those who have not been flooded. Information on flood risk is mainly acquired informally from relatives and friends or more formally from the local authority ( Appendix B; Table 40). National and local media are also significant sources, and Floodline is an important source (33.2%) but only for those who have been flooded. It comes as no surprise that respondents who have been flooded worry significantly more often than respondents who have not been flooded ( Appendix B; Table 41). The most important triggers for such worry are "heavy rainfall" or "seeing the river rise or in spate" (especially for those who have been flooded), followed by "forecasts of heavy rainfall" and "reports on flooding in the media" ( Appendix B; Table 42).

Managing flood risk

4.41 Contents insurance is held by 90.8% of respondents, this increasing to 94.7% for those who have been flooded, a significant difference between respondents ( Appendix B; Table 43). However, when asked whether the cover specifically included flooding ( Appendix B; Table 44), these values (still significantly different) reduced to 72.2% (all respondents) and 84.0% (flooded respondents). A significantly higher excess for contents insurance was reported by households that had been flooded: higher excesses being required for 27.4% of all respondents and 36.8% of those who had been flooded ( Appendix B; Table 45).

4.42 We also explored whether the take up of contents insurance differed according to housing tenure both at the time of the flood and afterwards at the time of the survey. At the time of the flood social tenants were much less likely to have contents insurance than those who either owned their property or were buying it via a mortgage (Table 4.10). By the time of the survey (1 to 13 years later depending on location) the take up had increased by 6.2% (council tenants) and 22.6% (housing association tenants) but had declined by 20.9% (private renting) possibly reflecting a high proportion of short-term tenancies. The increase for housing association tenants points to a marked change in behaviour, albeit by a very small sub-set of the whole population surveyed.

Table 4.10 Contents insurance at time of flood and at time of survey, by housing tenure

4.43 'Pay-with-rent' contents insurance is known about by 36.9% of council tenants, this value being slightly higher (39.1%), but not significantly so, for those who have been flooded ( Appendix B; Table 47). However, 73.9% of social tenants have contents insurance with 7 out of 87 council tenants having elected the 'pay-with rent' route (Table 4.11). Overall this means that around 75% of social tenants have contents insurance with around 10% using 'pay-with-rent' schemes, and around 25% with no insurance. When asked why contents insurance had not been taken out, 46.9% of social tenants cited cost and 21.9% the non-availability of cover ( Appendix B; Table 49). Whether or not the c.25% of social tenants without cover can be persuaded to join 'pay with rent' schemes will depend, in part, on insurance companies making cover available and the affordability of the premiums.

4.44 Householders' take up of flood alleviation measures is strikingly higher for those who have been flooded (60.7%) compared with those not flooded (31.1%). The most favoured responses are not having sentimental or irreplaceable items downstairs, the purchase of removable flood guards or sandbags and keeping the drains and ditches around the property cleaned out (Table 4.12). Receiving demountable flood guards or sand bags from the local authority is also a popular option, alongside building protective walls.

Table 4.11 Uptake of 'Pay-with-Rent' schemes among council tenants, by whether flooded (per cent of respondents)

Chi-squared = 0.925; df=2; p>0.01

Table 4.12 Flood alleviation measures taken, by whether flooded (per cent of respondents)

4.45 Respondents who have not been flooded register similar priorities, but the percentage take up is often very small. Reasons given for not undertaking individual flood alleviation measures were mainly reliance on others to build flood defences or improve drainage (by implication this was a local authority duty), followed by an expectation that insurance would cover losses and the perceived low likelihood of a flood of similar severity (Appendix B; Table 51).

4.46 Finally, we explore flood risk management based around warning systems. Floodline was known about by 74.2% of those flooded but by only 59.4% of those not flooded (a significant difference - see Table 4.13). Usage of Floodline is also significantly higher amongst those who have been flooded (35.5%) compared with those not flooded (13.1%, Table 4.14). High levels of customer satisfaction with Floodline are reported across flooded and non-flooded households - 77.7% and 79.3% respectively finding the service "helpful or very helpful" ( Appendix B; Table 54).

4.47 Of respondents who had been flooded, 51% received a formal flood warning (official phoning or knocking on the door, loud hailer, automatic telephone messaging), this significantly falling to 34.4% for those not flooded ( Appendix B; Table 55). Just over half of the respondents (52.7%) received one flood warning, this falling to 27.8% for two warnings and 19.5% for three warnings ( Appendix B; Table 56), these values being similar for flooded and non-flooded respondents.

Table 4.13 Awareness of Floodline, by whether flooded (per cent of respondents)

Chi-squared = 28.369; df=1; p<0.001

Table 4.14 Use of Floodline, by whether flooded (per cent of respondents)

Chi-squared = 67.483; df=1; p<0.001

4.48 In terms of future warnings (Table 4.15), it is striking that only 46.2% of households that had been flooded and warned were "confident or very confident" of a sufficient warning in the future, this falling to 21.2% for those not warned and flooded. By contrast, those who had not been flooded took a significantly more positive view with 60.7% of those warned and 40.2% of those not warned being "confident or very confident" in future warnings. Prior experience of being warned raises confidence but, paradoxically, to lower levels amongst those flooded than those not flooded. On the other hand, those who have been flooded but not warned are nearly half as confident as those who have been neither flooded nor warned. Thus having received a warning in the past increases confidence in getting sufficient warning next time, but this is lessened by the actual experience of being flooded, possible reflecting what can be done with "sufficient" time. Surprisingly, those who have been warned but not flooded, have by far the highest confidence in future warnings; but this may not be sustained through repeated false alarms.

4.49 However, prior experience of being flooded did result in significantly higher levels of confidence in knowing what to do in a future flood (65.3% "confident or very confident" - Table 4.16) when compared with those who had not been flooded (52.7% "confident or very confident"). The most preferred means of being warned were an official knocking at the door, individual action (checking the river) and a loudhailer in the street, television or a phone call ( Appendix B; Table 59), with a mix of these actions being favoured by a large number of respondents.

Table 4.15 Confidence of sufficient warning of a future flood, by whether ever received a warning and whether flooded (per cent of respondents)

*Chi-squared = 49.637; df=3; p<0.001
**Chi-squared = 30.262; df=3; p<0.001

Table 4.16 Confidence of knowing what to do in a future flood, by whether flooded (per cent of respondents)

Chi-squared = 26.401; df=3; p<0.001

4.50 We now explore whether receipt of a flood warning varies with socio-economic status. Neither age, nor income group, nor occupational class nor housing tenure register any significant difference in terms of flood warning ( Appendix B; Tables 60-63). This implies that the agencies responsible for dissemination do so in an even-handed manner to the 42% of sampled households which received a warning.

4.51 Finally we examine the acceptability and responsibility of a range of flood management policies. The most acceptable forms of flood management (considering effectiveness, cost and fairness) are structural flood defences (favoured by 91.1% of all respondents), closely followed by fitting valves to sewers and storm drains (89.0%), a flood warning service (88.5%) and using upstream storage in reservoirs (86.0%). Neither demolition of the most threatened properties with compensation or relocation of residents (39.2%) nor payment to rural land owners to increase storage of water in the soil (72.1%) approach the levels of support accorded structural flood defences (Table 4.17). But respondents who have been flooded are more supportive of demolishing threatened properties than those who have not been flooded (45.1% compared with 33.3%). Overall, structural measures are still seen as providing the first line in flood defence.

Table 4.17 Acceptability of flood management policies, by whether flooded (per cent of respondents indicating policy is acceptable)

4.52 When respondents were asked where they thought the main responsibility for protecting property from flooding should lie, responsibility was unequivocally assigned to local authorities by 58.4% of respondents, and to a lesser extent to the Scottish Executive (31.9%), with fewer than 20% of respondents identifying SEPA and Scottish Water (Table 4.18). When these answers are sub-divided according to socio-economic characteristics, households with septuagenarians preferentially allocate responsibility to the Scottish Executive and away from local authorities ( Appendix B; Table 66), whereas households with an annual income of less than £20,000 reverse this allocation with local authorities overwhelmingly favoured ( Appendix B; Table 67). When the same question is asked across occupational classes, the local authority continues to be most favoured responsible body, but by a lower percentage of those in professional and managerial grades (Appendix B; Table 68). Unsurprisingly, council and housing association tenants overwhelming see flood protection as a local authority duty. However, this declines sharply in favour of the Scottish Executive amongst home-owners and households with private landlords ( Appendix B; Table 69). When asked where responsibility for flood protection actually lies, 22.9% of respondents accepted individual responsibility (up from 8.3% in terms of where responsibility should lie), with local authorities, the Scottish Executive, SEPA and Scottish Water identified as responsible authorities as before, but by a lower proportions of respondents (Table 4.19).

Table 4.18 Views of where main responsibility for flood protection SHOULD lie

Table 4.19 Views of where main responsibility for flood protection DOES lie

Table 4.20 Willingness to pay extra Council Tax per annum for flood protection, by whether flooded (per cent of respondents)

4.53 Finally respondents were invited to specify their willingness to pay additional council tax which would be hypothecated for flood protection measures (Table 4.20). More than half of respondents were unwilling to make any additional payment. Of the remainder 8.5% were prepared to pay £100 or more, 6.4% £50-99 and 15.3% up to £49. Those who have been flooded are more willing to make a higher contribution - 10.2% of those flooded would pay more than £100 compared with 6.6% of those not flooded. Income is also a statistically significant discriminator of willingness to pay ( Appendix A; Table 72) with 15.9 % of higher income households (>£20,000 per year) offering to pay £100 or more, compared with only 7.0% for lower income households (<£20,000 per year). When analysed by housing tenure, those owning their home with a mortgage express the greatest willingness to pay (albeit with nearly half not willing to pay any extra), while social tenants express the least willingness with almost two-thirds not willing to pay any extra council tax for flood protection ( Appendix A: Table 73). Again this is a statistically significant discrimination.

Summary

4.54 This chapter has reported the findings from a questionnaire survey of households in eight locations flooded within the period 1993-2005. Overall, roughly equal numbers of flooded households and households at risk, but not flooded, were surveyed. The single most important finding is that intangible impacts are more severe than tangible ones. Also immediate impacts are more severe than lasting impacts, but the latter (which include long-term health effects) can individually be very onerous, especially for vulnerable households where modest incomes may inhibit post flood recovery. Flood warnings and insurance can help mitigate flood impacts. But only 33% of those who have been flooded use Floodline, and contents insurance take up is significantly lower for social tenants than owner-occupiers. Only 13% of respondents correctly identified owners or landlords as mainly responsible for flood defence.

« Previous | Contents | Next »