Bulletin Board
IMAGE Why The Finance Foundation?

The Finance Foundation is an independent think tank that aims to encourage informed debate about financial services, explain what the sector does, and suggest ways it could work more...

IMAGE The Future of Cash

We are actively engaged in the current debate on the future of cash. While recognising the drivers towards digital banking, we are concerned that insufficient attention is being given to the needs of...

IMAGE Supporting Growth

Debate continues about the importance of small and medium-sized businesses for growth and banks' willingness to lend to them. Director Andrew Freeman's report, 'Finance for Growth, Challenging...

IMAGE Rebuilding Trust

Corporate culture and behaviour in the financial sector have been under scrutiny since the financial crisis as never before. Shortly after the Parliamentary Commission for Banking Standards published...

“When I’m 84”. Locking the Door on the Older Old: The Challenge Facing Britain’s Banks.


Methodological Note

Summary and Overview

Our three core objectives in conducting this research were:

  • To investigate how people aged 80 and over manage their routine financial tasks such as paying bills, taking out cash and keeping track of money and to include the perspective of those who provide informal care and support to them;
  • To conduct original research – collecting both qualitative and quantitative data from the older old and their carers – on the type of support older people need and the nature and extent of the problems and barriers they encounter, whether this be due to physical and sensory impairment, poor capability, exclusion from mobile and digital access or more general vulnerability to poor conduct or even fraud;
  • To draw out key findings and make recommendations for the industry and policy makers on the changes that need to be made to support future cohorts of the very old, taking account of how these needs may change over time.

This note provides additional technical detail on the two surveys – one of older people aged 80+ and one of informal carers supporting people of this age.

Survey of older people

1. The market research company Opinium Research was commissioned to interview a representative sample of people aged 80 and over across Great Britain. Fieldwork took place in October/November 2015, and followed earlier focus group work in which issues were explored and tested. A brief pilot of the interview with around 10 respondents took place prior to the full fieldwork, leading to some minor adjustments. 

2. Around a quarter of the 175 interviews in the full survey were conducted with people in the street, with the remainder being carried out by interviewers who called at people’s houses in selected areas of the country. All the normal assurances around confidentiality, anonymity and professional independence were given, and participants were told it would be a few questions on ‘how you manage your day to day finances’, taking around 10 minutes, with nothing specific about amounts of money etc. Interviewers reported high levels of interest and co-operation from older people. Full detail of the geographical, age and gender spread achieved is given in the annex. The required outcome in terms of age and gender representation was achieved without imposing quotas, and the regional spread was good.

3. Actions to achieve a representative sample that avoided either over-sampling the most or the least able in the age group:

  • The most important decision was to use face to face interviews in order to be as inclusive of the age group as possible, given the known limited use of the internet by people aged 80 and over.[1]
  • To avoid the alternative risk of including those who were not fully managing their own affairs, or were otherwise less capable, we put in a number of screening conditions. First of all  interviewers were instructed to assess whether a potential interviewee would be capable of answering the questions, and in particular to stop any interview if there was any sign or mention of dementia.

People were also screened out if:

  • They lived in a retirement home or
  • They said when asked at the start of the interview that someone else managed most of their finances on their behalf, either informally or via a Lasting Power of Attorney.

4. The sample achieved therefore includes people who were either solely in charge of managing their own finances (58%) or said they managed their finances jointly with somebody else such as their partner or family member (42%). Some - 54% - said that family members or friends sometimes helped them with day to day financial tasks such as getting out cash or finding better deals, even though they were still in overall control. 

5. Around three-quarters of the sample were living independently in their own home, as follows:

Living on own in own house: 62%

Living with husband/wife (or partner) in own house: 15%

Sheltered accommodation: 14%

Living with another family member: 9%

This is consistent with ONS data on the high proportion of the older old who live alone, and suggests that this is not an unrepresentatively dependent group.

6. Although we did not ask this interview group directly about their health problems (in order to avoid extending the interview and potentially introducing long digressions on subjective subjects) we did establish that 13% were in receipt of attendance allowance, with this being greater among those aged 85 and over, as would be expected. By contrast 33% of older people in the carers sample received attendance allowance, even though they were just as likely to be living in their own house as the interview sample (see details in annex).

7. In terms of digital capability, the interview sample appears representative. Ofcom data quoted in the report and taken from their 2015 Metrics Bulletin (p.9 and p.11) suggest that approximately 11% of people aged 75 and over are using internet banking (compared with a third of 65-74 year olds and nearly half of 55-64 year olds). In our somewhat older sample of people aged 80 and over we found that 9% were engaged in internet banking (13% had said they used the internet for either shopping or banking, but only 9% for banking). This appears pretty much as expected based on the age-related decline in internet banking within the Ofcom data.

8. Data collected on sources of income again suggest that the sample is not skewed towards people who are more likely to be financially excluded. In addition to the expected high receipt of state pension, around half were in receipt of an occupational pension (either their own or from their spouse) and two-fifths said they had income that came from investments or savings. A further 10% declined to give details of income source. (Further detail on this can be found in the Annex).

Survey of informal carers

9. The online panel survey of carers was designed to capture the issues that older people and their carers have to manage where an older person is getting support in managing day to day financial matters. In many cases this was quite light touch support/advice on getting the best deals or getting a lift to the bank/cash point, but in other cases there was much more hands-on involvement with basic tasks.

10. The sample was nevertheless designed to keep a focus on older people who still had control of their affairs – it therefore excluded people with dementia (unless it was only at the initial diagnosis stage), anyone whose finances were being managed via a full legal arrangement such as an LPA (although some with third party mandates were included) and people who were living in residential accommodation.

11. The carers sample – 255 individuals in total - was recruited from Opinium’s existing online panel and was conducted electronically. In order to be included panel respondents were asked if they “sometimes help an elderly parent, other relative or friend aged 80 or over with managing some of their day to day tasks”. This was the key screening definition for ‘informal carers’ that was used with the intention of capturing family and similar support as distinct from support within the formal care sector (although clearly some of those receiving informal family/other care would be likely to also have structured formal care in place).

12.The next step was to switch from “day to day tasks” to ask how far these self-identified informal carers give “practical help or advice on managing day to day finances e.g. paying bills, taking out cash, keeping track of money or savings etc.” Only those who said either that they “often” or “sometimes” give such support were then included. (529 people who had identified under the broader definition of informal carers were excluded as they said they either rarely or never gave help with day to day finances).

13. Although they may overall be further along the spectrum of dependency than the older people interviewed in the face to face survey, some of whom are still managing well without help, we did not want to include people whose relatives (or similar) had actually taken full control via an active Lasting Power of Attorney (LPA). We therefore screened out any cases where a respondent had said they sometimes or often gave help with financial matters which included them or another person exercising LPA – 1019 respondents were excluded as a result of this condition.

14. The screening questions allowed for the fact that an informal carer might be helping more than one older person, enabling them to stay in the survey if at least one of those did not include an active LPA and then asking them to focus on the “main” person to whom they give help for the rest of the survey. (One third of the sample was helping more than one person with day to day financial matters – in many cases, given the sample characteristics, this might be where an adult child was providing help to both parents. Interestingly, in this context carers aged 20-39 were much more likely to be supporting more than one person – 51% compared with the sample average of 34%, perhaps reflecting the greater likelihood in this age group of having both parents alive.

15. Further screening questions were used make sure that those remaining did not include anyone providing support to someone under the age of 80 and to limit the carers sample to those aged under 75 (this was mainly so that we retained a focus on carers who, being younger, were less likely to be experiencing the same problems with financial access as those they were helping; we also wanted to avoid picking up what would probably be a rather unusual group of over 75s – those actively using the internet as part of an online panel. We also excluded anyone who was the spouse of the person they gave support to (this was to ensure we were genuinely picking up an informal care relationship rather than a situation in which one person within a couple was just helping a less able/interested partner). Initially 7 people aged over 75 were excluded and subsequently 9 who were the spouse of the person they cared for were screened out. The fact that these numbers are so small confirms our view that in practice very few of the oldest people are taking part in the online panel when compared with younger age groups. This is further supported by the fact that in the 70-74 age group, which we did include (providing the carer was not a spouse) we only had 9 respondents.

16. Finally we excluded anyone who was providing informal support to an older person who was living in an assisted/retirement home, where it would be expected that their need to exercise control of day to day finances would be less, or who had dementia/ Alzheimers that had gone beyond the stage of just slight impairment (minor memory issues) to something more disruptive. This excluded a further 25 and 12 respondents respectively. As older people at this level of dependency/incapacity are highly likely to have an active LPA we assume that a significant proportion of those screened out earlier for having an LPA in use (1019 in total) would fall into one of these categories of dependency and so would have already been removed.


Sample Characteristics – Further detail

Older People’s Sample

17. 175 face to face interviews were conducted mainly in urban or suburban centres designed to give a spread across the North (35%), Midlands (25%) and South (40%) – achieved figures in brackets. Within these figures, Wales and Scotland were well-represented – 6% of interviews being in Swansea and 7% in Glasgow. Interviews were concentrated in areas where there was a good representation of older people and included: Croydon (7%), Harwich (6%) Romford (7%) Letchworth (7%), Norwood (5%), Dover (7%) Swansea (6%), Birmingham (7%), Nottingham (7%), Stourbridge (6%), Hull (7%), Durham (7%), Glasgow (7%), Manchester (7%), Rochdale (7%). The lack of coverage of the south coast and south west of England to some extent reflects the decision not to interview in rural areas (although they were included in the carers’ sample – see below).

18. The interviewers were asked to operate within the following quotas: a ratio of 60:40 females to males (to reflect the average gender balance in the age group; at least 30% of interviews to be with over 85s, and at least two-thirds of interviews to be in home.

19. The outcome was 61% of interviews with females, and 39% with males, 74% conducted in the interviewees’ home and 26% in the street and 49% being 85 and over – all of this falling out naturally without the quotas being enforced. For both genders the age balance was similar – more or less half were below 85 in each case – (50.7% males and 50.9% females). However females 85 and over were rather more spread into older ages – with 9.4% of all females interviewed being 90 or over compared with 7.2% of males, entirely consistent with population spread.

20. There was very little variation in the demographic characteristics of those interviewed in the street rather than at home, with regional variations small and just a slight tendency, probably not statistically significant, for males aged 80-84 to be more likely to have an in street interview (31% of the interviews with this group, compared with sample average of 26%).

Carers’ sample – the carers

21. 48% of the carers in the sample were male and 52% female with these ratios remaining broadly constant across all ages of carers. Female carers were, however, more likely to be giving financial support to more than one person - 38% compared to 27% of males, although this is much less strong than the age effect – where 51% of 20-39 year olds support more than one person (compared with 27% in all other age groups).

22. Carers were spread across the age distribution, although, as would be expected concentrated amongst later middle age: 6% of our sample were aged 20-29, 16% aged 30-39, 18% aged 40-49, 33% aged 50-59, 22% aged 60-69 and 4% 70-74.

23. In response to an optional question on ethnicity, 85% of carers identified themselves as white British, 5% as white other and 4% as from an Indian/Pakistani or Chinese background. None were from African/Caribbean backgrounds, the remaining part of the sample either not stating their ethnicity (3%) or being from mixed or other minority backgrounds (2%). The overlap with the ethnicity of the person they cared for was very strong – 98% of white British carers were caring for a white British older person (99% if Irish older people included), and for those from an Indian and Pakistani background the overlap was 100%. This is consistent with most support being ‘within family’ support. There was no significant gender difference amongst different ethnic groups. In each ethnic group females were around 60% of those being cared for and carers were roughly half male and half female.

Carers’ sample – the older people being cared for

24. Of those being cared for, 40% were males and 60% female, exactly the same as for the interview sample and reflecting the overall population. 58% of those being supported were aged 80-84 and 42% aged 85 or over, giving a slightly increased emphasis on the younger old than the interview survey. However amongst the 85 and over, fully 16% of the total sample was 90 or over, a greater proportion than the interview sample, where the equivalent figure was 9%.

25. While the interviewees had been sampled in urban or suburban areas the online sample was not geographically constrained. 38% of the older people identified from the carers survey were from cities or towns, 47% from suburban areas and 15% from rural areas. The sample included people across all regions, with the largest numbers recorded in London (18%), South East (14%) and North West (14%) broadly consistent with overall population numbers (although London is somewhat over-represented presumably driven by the characteristics of the participants in the online panel).

Relationship to Carer

26. Carers providing informal financial support were asked to indicate the nature of their relationship to the main person they were supporting. Half - 49% - were helping a parent, 27% another family member – including aunt or uncle (7%), sibling (1%), and other relative (18%) - and 24% were giving help to a friend (11%) or neighbour (13%).

27. These figures varied markedly in relation to the age of the carer. Only 25% of those aged 20-39 were giving support to a parent, with 38% supporting other relatives (of which grandparents are probably a significant group) and 38% supporting older friends or neighbours. For carers aged 40-49 the emphasis was on parents – 66% of those being supported – and much less on older friends or neighbours – accounting for just 10% of those in this age group. However for those aged 60-74 friends and neighbours become a bigger focus of support – 40% - than parents – 38%, still significant nevertheless.

28. Looking at other carer characteristics, there was no significant gender difference in the relationship that they had to the carer, although there were differences by ethnic group – those of Asian origin being more likely than average to be supporting a parent (55%) or other relative (36%) than a friend or neighbour (9%).

29. Some variation was also apparent in the characteristics of those being cared for. Older people from social classes ABC1 were less likely to be parents of their carer - 46% - and more likely to be friends and neighbours - 40% - than those from social classes C2DE where 57% were parents and 17% friends or neighbours. Likewise in urban areas a slightly smaller proportion were parents – 44% - and a larger proportion friends or neighbours – 30%. Amongst those living on their own in their own house – at three-fifths of the sample the single largest group - the person being given help was significantly less likely to be a parent – 37% - and more likely to be a friend or neighbour – 29%.

30. The two groups most likely to be the parents of those giving help with finances were those living on their own as part of a couple (70%) and those actually living with the carer (86%). As far as the gender of the person being supported is concerned, females being cared for were just as likely to be parents (around half) but more likely than average to be other relatives – 31% – than friends or neighbours – 20%. For males being cared for the picture is a mirror image – while half are parents, 21% are other relatives and 30% friends or neighbours.

Level of Support


31. All of those in the carers’ sample were providing some kind of support with day to day finances to an older person aged 80 or over as a criterion for inclusion – around 58% of those providing this support said it was something they did “sometimes” with the balance of 42% saying they did it “often”. The percentage of older people who were often being given help was the same on average for both males and females. However whereas for males this did not rise much with age, with 41% of 80-84 year old males often receiving help and 44% of those aged 85 and over often being helped, the gradient was steeper for females – with fewer of the younger 80-84 year olds – 35% - often receiving help but proportionately more of the older females aged 85 or over – 51% - being helped often.


32. Although we excluded anyone who was exercising Lasting Power of Attorney for an older relative or friend, this still left open the possibility that other arrangements short of LPA might be in place. We asked if the respondent or any other person either had a third party mandate, or a joint account (either joint or separate signatures) in place. 28% of respondents recorded that at least one of these arrangements was in place, with this being rather higher for males – 33% - than females – 24%.  For males, the control arrangement was a third party mandate in more than half of cases, whereas for females third party mandates accounted for only a third of the formal processes set up, joint accounts with both signatures required being the more common arrangement. There was no discernible difference in the likelihood of having an arrangement of the above sort in place according to the age of the older person.


33. Carers were asked whether they lived near enough to the older person they helped to be able to easily drop in on them. 83% confirmed that they did, with no significant variation by gender of the older person cared for and interestingly no marked difference according to the social class of the older person or whether they lived in a rural or urban area. What did emerge though was a tendency for those caring for the very oldest age groups to be less likely to be easily able to drop in on them – 77% of carers of those aged 85 and over could easily drop in compared with 88% of the carers of 80-84 year olds. However, this is perhaps less surprising when put alongside a question asking if the respondent was the only or main carer. While on average 65% said that they were the sole or main carer this was rather less for those looking after a person aged 85 and over, so it may be that in these cases one of the other people also providing care is in a better position to drop in on them.


Comparison of older people’s and carers’ samples

34. We have seen above that the gender and age balance of the two samples of older people (from the interview survey and the older people linked to the carers in the online survey) was similar, with just a small difference in geographical spread mostly relating to the inclusion of older people in rural areas in the carers sample (15% of the sample).

35. Interestingly the pattern of living arrangements and the reported sources of income for older people from the carers sample was also very similar to that of the interview sample of older people (with the exception of attendance allowance) again suggesting that the two independently-generated samples produced a similar demographic of older

Living Arrangement

36. As can be seen from the table below there is a very strong overlap between the two surveys in terms of living arrangements, with roughly three quarters living in their own home, and by far the majority – around three fifths of all those included – living on their own. Within this overall spread, there is a small gender difference, with males a little more likely than females to be living in their own house but also more often living with their wife/partner (likely to reflect the fact that men are more likely to have a surviving spouse than the other way round). The slightly lower proportion of females from the carers sample who are living independently is likely to reflect the rather older age distribution of females in this sample – 18% being 90 or over.


                                                              All                              Males                      Females

                                                       A              B                  A              B                 A            B                                               

In own house – on own               62%         59%              57%        53%             66%        63%

In own house – with partner       15%         17%              22%         27%             11%        10%

Total living in own house         77%         76%              79%        80%              77%       73%

Living with other family                 9%        13%                4%         12%             11%        13%

Sheltered Accommodation         14%         11%             17%           8%             11%        14%  

Total in supported living          23%         24%             21%         20%             23%        27%

Note 1 – A is the interviewee sample and B is older people from the carers online survey


Occupation/income and receipt of attendance allowance

37. In the carers’ sample, respondents were asked about the occupation of the person they helped. 62% were from social classes ABC1 and 38% from C2DE, according to their carer’s assessment of the main occupation of the chief income earner in their household when they were employed in the past.

38. Although the interviewee sample was not asked for their occupation when in work, they were asked about their sources of income.  As expected nearly all had income from a state pension – 86% - and 51% had some form of occupational pension either from their own work or their spouse’s. Around two fifths had money from either investments or savings while 13% were in receipt of attendance allowance (for themselves, not their spouse). This latter category was the only one that varied much by age, entirely in line with expectations, with 9% of the 80-84 year olds receiving attendance allowance and 16% of those 85 and over doing so.

Which of the following sources of income do you have?






Base: all respondents











State pension (including pensioner credit)











Attendance allowance (for yourself, not spouse)











Works/occupational pension (either own or from spouse)











Money from investments and / or savings











Not prepared to say












  1. 39. Carers were also asked the same question about the sources of income that the person they looked after received. The responses were almost identical in respect of state pension (86%), occupational pension (52%) and money from investment or savings (43%). Only with regard to receipt of attendance allowance was there any significant difference. Whereas 13% of the older interviewees were in receipt of attendance allowance, 33% of the older people being cared for received it, and this showed the same marked age variation as for the interview sample with 27% of the 80-84 year olds getting the allowance and 42% of those aged 85 and over. This confirms very clearly the gradation between the two samples of older people in terms of the extent of their dependency. Reading this alongside the data in the report on significant health problems reported by carers amplifies the issues for this group although it is in fact not inconsistent with other information on health limitations experienced by this age group.

Hilary Cooper

Senior Associate, The Finance Foundation

This email address is being protected from spambots. You need JavaScript enabled to view it.

[1] The report by the Social Market Foundation, Balancing Bricks and Clicks, also includes findings about the financial habits and views of those aged 75 and over, but as it is based on an online panel survey it is likely to have a high level of bias in favour of more affluent, better educated and healthier people in that age group – the majority of people in that age range do not use the internet and very few would be participants in an online panel.