3. Demographic and socio-economic structure of the population samples As mentioned in Section 1.2.1, classical indicators of vulnerability include demographic and socio-economic characteristics of a particular population. In the following sections, the samples in the surveys are compared for gender, age and in terms of household composition, socio-economic stratification and vulnerability factors including health where the variables are available. The survey samples have been compared with the Census data for England and Wales (available from www.statistics.gov.uk/Stat.Base) on a few of the variables where comparison was possible (Table 3.1). Because the samples were drawn from at risk or flooded populations rather than from general populations within particular areas, it is not possible to determine in most cases whether deviations from the national picture reflect differences in these local areas, or in populations in flood risk areas as compared with the nation overall or some selection or other bias or in the survey samples. Table 3.1 summarises the socio-demographic characteristics of the respondents and households where interviews took place for all the three surveys.
Table 3.1: Social structure and housing/residence characteristics of the samples
* Lower Thames contains two residents in the age group 16 -24
**% of males and females aged 16 and over
*** % of people aged 16 and over in households
Gender and age
The Warnings Survey and Lower Thames Survey samples were reasonably well balanced in terms of the gender of the respondents (Table 3.1). However, this was not the case for the Intangibles Survey, both those at risk and the flooded. The 2001 Census shows that 52% of the population in England and Wales (over 16) are female. In the Intangibles Surveys, 61% of flooded sample respondents were female and for the at risk sample the proportion was 57%. The imbalance may be the result of interviewer selection bias and of interviews taking place on weekdays, rather than at weekends or evenings when women tend to be more available than men.
The flooded and at risk samples in the Intangibles Survey were drawn from the same locations. However, they differed significantly in various ways. Indeed the at risk sample stood out as different from all the other samples in terms of age. The at risk sample was younger than the flooded sample with more respondents in the 18-34 age group and fewer people aged 65 and over. The at risk sample was thus closer to the population of England and Wales than the other samples. The mean ages for the at risk and flooded sample respondents were 45.4 and 54.5 respectively which can be compared with the Census 2001 figure of 47.5 for adults aged 18 and over in England and Wales. All the survey respondents apart from the at risk were older than the population of England and Wales as indicated by the Census data, with fewer residents in the under 45 age groups and more in the 45-64 age group, the 65 and over and indeed the 75 and over age groups. It is only possible to speculate on why this is the case. It is possible that those living in flats above ground floor level who were likely to be younger people were under-represented because they would not have been flooded above floor level and would have been excluded on those grounds. Older people are more likely to be housed in ground floor flats and bungalows and may be over-represented among those flooded as a result. It is also possible that locations near rivers were seen as attractive places to live and therefore residents were less likely to move away from them as they grew older.
The at risk sample in the Intangibles Survey also differed from the other samples in terms of the households represented, reflecting the differences in respondents’ age. The at risk sample contained more households with young children under ten years of age and fewer households containing elderly members, both those aged 65-74 and more particularly those aged 75 and over than the other samples (Table 3.1). One might expect that households with young children would be particularly vulnerable during flood events. With older children and young people under 18, it might be different since teenagers might be able to help with the care of young children and with taking action to protect property. In the same way households that included older people, particularly the very elderly aged 75 and over, might also be expected to be handicapped as compared with other households.
The size of the households was compared for the Intangibles Survey and the Warnings Survey. Data on household size were not included in the Lower Thames Survey. All the survey samples contained significant minorities of one person households (Figure 3.1). However, the at risk sample included many more people under 65 living on their own, while in the other samples, older people aged 65 and over predominated among the single person households. It can be hypothesised that those living alone will be more vulnerable in the event of flooding because they will be without others in the household to consult on what to do and to help take action to prepare for flooding. Older people living alone might be expected to be particularly vulnerable.
The most common size of household across all the samples was a household consisting of just two people reflecting the age and stage in the household life cycle of the residents in the flood risk areas. There were very few larger households containing five or more residents of all ages including children as well as adults (Table 3. 2)
Table 3.2 Household size: Intangibles Survey and Warnings Survey
Flooded % (N= 980)
Five or more
3,3 Socio-economic stratification3.3.1 Social grade classification
Socio-economic status was highlighted in Section 1.2.1 as a possible indicator of social equality, and therefore social vulnerability. The social grade classification of respondents used in the surveys is that used by market research companies (The Market Research Society, 2002). This social grading is based upon the occupation of the ‘Head of Household’ or ‘Chief Income Earner’ (CIE) in a household. In the surveys analysed for this report, the CIE provided the basis. The surveys excluded residents in communal institutions and were of residents in ‘private households’ i.e. people living alone or in a group together whose food and household expenses are managed as one unit. The CIE is the person in the household with the largest income, whether from employment, pensions, state benefits, investments or any other sources regardless of gender. Where two people in a household have equal incomes, the older person is taken as the CIE. The CIE is graded according to his or her current occupation. An indication of the occupations and their grading is given in Table 3.3.
Where the CIE is retired, the grading is according to the previous occupation if the person has an occupational pension, a state earnings related pension or private means i.e. they are not dependent on the basic state pension and other state benefits for income. Widowed, divorced or separated persons are graded according to their own occupations unless they do not work or receive a pension or maintenance from their former spouse. Occupations are listed by rank and the appropriate grading attached to the rank is indicated.