Ageing in the middle

The idea of the ‘squeezed middle’ is at the heart of today’s political debates about economic recovery and living standards. It’s a pithy phrase reflecting evidence that people in the middle of the income distribution have seen their incomes stagnate while the costs of living keeps rising. But almost all the recent discussion and analysis has focused on people ‘in the middle’ who have not yet reached retirement. So what about older people with mid-incomes? Are they squeezed too – and indeed what else can we say about their lives?

With support from Hanover, the Fabian Society carried out a short project to try and answer these questions. We wanted to focus on older people ‘in the middle’ because we felt this group was being overlooked in political and media debates about later life. Perspectives on ageing are dominated by how people think about the ‘top’ and the ‘bottom’. There are two competing narratives of later life: the wealthy baby-boomers with a surfeit of money and leisure; and pensioners on the breadline facing poverty, isolation and ill-health. The truth is that the majority of older people today are somewhere in between – neither rich nor poor – and the middle is expanding as a result of recent successes in reducing pensioner poverty.

We also asked what changing experiences of ‘ageing in the middle’ should mean for public policy and the design of services. Public provision plays a critical role in securing wellbeing in later life but there is little objective and explicit debate on the right balance of support for older people ‘in the middle’. Major entitlements continue to be available universally, particularly the NHS and state pension. But others, including state-supported housing and social care, have been targeted quite narrowly according to need. With regard to social care, the government has accepted the Dilnot Commission’s recommendations to widen access to public funding in a set of proposals designed to help people in the middle particularly; but there has not been similar debate on housing support for mid-income older people. There is also controversy regarding the a range of age-related universal benefits such as Winter Fuel Payment, which have not been affected by spending cuts in the same way as entitlements for younger households.

This all begs the question of when and why public provision should be offered universally, setting aside essential help for people with low incomes – and how independent providers should respond when publicly-funded support is not available. These issues touch on questions of intergenerational fairness and age equality. With old age an increasingly poor proxy for income, are there still good reasons to differentiate according to age in the design of services, tax and social entitlements? When are age-based eligibility criteria appropriate? And how should expectations and service offerings change when so many older people are neither rich nor poor?

The project was a quantitative study, which included a review of existing published data-sources and original secondary analysis of wave 4 of the English Longitudinal Study of Ageing (ELSA). We looked at the incomes of older people ‘in their middle’, but also their wealth, housing conditions and social circumstances. The full results are published in an accompanying Fabian Society report Ageing in the Middle.

For most of our work we considered people with middle incomes, although we also briefly looked at people with middle wealth (we found there was only a partial overlap between these two groups, but little difference in their social characteristics). We wanted to profile the lives of people in the middle of their own generational cohort, so we considered the middle fifth of specific age-groups rather than the population as a whole. With most of the published studies this meant looking at the middle quintile of older people, but in our own ELSA analysis we looked in more detail at the middle fifth of each ten year age-band. Incomes were defined and measured differently in each study we reviewed, but the impression they all give is that older people ‘in the middle’ are some way above the poverty line, but far from being affluent compared to the norms of British society.  For example, two official studies found that in 2010/11 the median income for an older household was £290 a week or £15,000 a year.

Middle incomes in later life remain significantly lower than those of people of working age, so in this sense older people in the middle are squeezed. Clearly many are finding budgeting very tough as living costs increase, particularly if part of their income comes from interest on savings. But that’s not the whole story. We found that over recent decades, mid-incomes for older people have been slowly catching-up with those for pre-retirement households and that they have continued to rise in the last decade, in contrast to the stagnation of middle incomes for everyone else. Current projections suggest this divergence will continue in the next few years. So although mid-income older people live on fairly modest means, our study casts doubt on whether they should be thought of as part of the ‘squeezed middle’ as the term is widely used.

As a consequence it would seem sensible to review all policies which appear to give special advantages to older people as a category.  With both the long-term convergence of incomes and the immediate living standards crisis for younger households, it is time to adopt a ‘presumption of equality’ in how public policy and services treat different age groups. That does not mean scrapping all age-based entitlements, for a generous state pension remains an essential building-block for a good old age. But it does mean reviewing eligibility rules on access to services, social security and the design of taxation.

Money is not everything however. Our analysis of the health and social data from ELSA showed that many mid-income older people face significant disadvantages, despite their financial cushion. Most strikingly, almost half the group had a longstanding illness or disability which limited their daily activity. One third reported being isolated or lonely, demonstrating that exclusion ‘behind closed doors’ is not just a problem for low-income groups. And one quarter reported problems with their accommodation, including noise, poor conditions and cold. Indeed we found that, when it comes to housing problems, people’s income seems to make little difference.

This shows that many mid-income older people continue to have significant social needs, despite their rising incomes. This is a key finding from the study which needs to be set alongside the positive news we found when looking from a purely financial view. Our analysis confirms that money alone is not sufficient to prevent disadvantage and policy makers and service providers must focus on overcoming the barriers associated with old age for people of all incomes. For some services this will take a change of culture and business model, as we explore later.

But in spite of the barriers, we should not forget that people ageing in the middle are very much part of mainstream British life. True, many have housing problems, ill-health or feel isolated – and at the time of the ELSA survey only three in ten were using the internet. But large majorities own their own home, enjoyed a holiday in the last year, have use of a car and are members of clubs and organisations. They also lead relatively healthy lifestyles, at least when compared to the contemporary British norm. Some older people ‘in the middle’ face real social pressures, but most share in the benefits of contemporary British life, and policy and service provision should not assume otherwise. That too is what we mean by a ‘presumption of equality’.

Incomes ‘in the middle’

Rising middle incomes for older people are in part a consequence of ‘cohort’ effects, in that people reaching old age have higher average incomes than the people they replace. This goes hand-in-hand with considerable differences between the middle incomes of people in early and late old age, with the median income of a 65 to 69 year-old one quarter higher than of someone aged over 80. Most of this is explained by cohort effects, although we found that a quarter of the fall is due to the greatly reduced incidence of paid work as people grow older.

The disparity between the middle incomes of different older cohorts is however nothing compared to the shift in middle incomes that has occurred over time. Since 1979 real incomes for retired people ‘in the middle’ have risen over one-and-a-half times from £7,000 to £18,500 (calculated on the basis of a two-person household). Half that improvement happened during Labour’s thirteen years in power.

Over the same period there’s also been a marked change in income inequalities among retired households. Inequality remains lower among older people than younger age-groups but it increased significantly in the 1980s and early 1990s, with middle income groups drawing away from poorer pensioners while the rich raced away still further. Then during the new Labour period the pattern changed. The gap between mid and low income groups stabilised, as both benefited from more generous state pension policies, and inequality between the middle and the top reduced a little.

The recent stabilising of inequality and rapid improvement in middle incomes both demonstrate the critical role of government pension policies for older people ‘in the middle’. Indeed, with around two-thirds of middle incomes coming from the state this should be no surprise. The last government’s policies not only reduced poverty in old age but gave a boost to middle incomes. Similarly the coalition’s proposal for a flat-rate single state pension is projected to lead to a noticeable increase to middle incomes for future pensioners.

The data from the last 30 years also shows a dramatic shift vis-à-vis younger age-groups, with mid-income retired households catching up with those who’ve not retired. In 1979 middling households of working-age had average incomes almost twice as large as those of older people (93 per cent higher); today they have just over one third more (37 per cent higher). Of course this is something to welcome and celebrate, as part of the steady decline of pensioner poverty, but it has profound implications.

Older people ‘catching-up’ with everyone else was not problematic while middle incomes were rising across the board. Perhaps it is more so today with growth in median earnings at a standstill: in 2010/11 the real incomes of the middle fifth of all households were no greater than in 2003/04, but middle incomes for retired households were 13 per cent higher. Since the financial crisis this disparity has become even more stark: real middle incomes have fallen by 5 per cent overall, but they have risen 5 per cent for retired households. This divergence is set to continue, with older people’s incomes expected to continue rising while younger people’s remain flat for some years to come.

In a deep economic crisis these effects should come as little surprise, since the drivers of income growth are different during working life and retirement. Retirement incomes are the result of long-term public policy and of economic conditions over an entire working life, while incomes prior to pensionable age are dependent on short-term economic forces. The divergence should therefore be addressed first and foremost by improving incomes for middle-income people of working-age rather than wishing away the recent gains older people have experienced. But just the same, the differing fortunes of the retired and non-retired ‘middle’ is something that cannot be entirely ignored when politicians consider hard financial choices.

Wealth and housing

Around 80 per cent of the mid-income older people in our analysis of ELSA were homeowners. This has important implications for disposable incomes because mostly these people don’t have to pay rent or mortgages. Indeed when looking at disposable incomes after housing costs, the median pensioner couple is now in the top half of the UK income distribution and the median single male pensioner is are only just below the national midpoint. Single women still lag well back, but nevertheless one independent projection suggests that, after housing costs are taken into account, the median disposable income for all pensioners may overtake that for the whole population in the next decade or so.

As well as affecting disposable income housing tenure, of course, impacts on wealth. So it is of little surprise that assets are highest among age groups where homeownership is most widespread: median household wealth peaks for people aged 55 to 64, at £243,000, and then gradually falls in late old age. The latest evidence indicates this is a ‘cohort’ rather than an ‘ageing’ effect. People in late old age have less wealth because they were less affluent over their lives and are more likely to be renting not owning; but on average retired people do not ‘decumulate’ their financial and property wealth as they move through retirement (in contrast to what economists would expect to see).

There is not yet consistent data for tracking changes in wealth over time, but in the last twenty years we know that the share of homeowners aged over 65 has increased dramatically, from 58 per cent to 76 per cent of households.  In the last decade this has been accompanied by a dramatic fall in the share of people aged under 45 who are owner occupiers, with the median 25 to 34 year-old now renting rather than owing their home. Both these effects relate to the affordability of housing at the point when each generational cohort were likely to be buying for the first time: homes were relatively cheap in the 1980s and 1990s but today the average first home costs four times a typical salary, even after the financial crisis. So while the rise in homeownership among older people has spanned many decades, the collapse among younger age-groups is mainly a recent phenomenon.

So what does this mean for intergenerational fairness and opportunity? It’s worth recalling that people with mid-incomes in early adulthood today have much higher incomes and general standards of living than older cohorts did at the same age. Nevertheless for today’s under-45s homeownership is far less affordable than it was for their predecessors; and if house prices remain at current levels relative to earnings, low levels of homeownership (and therefore personal wealth) will become a permanent characteristic of this generational cohort not an effect of their youth or the economic cycle. As older homeowners die, property will pass increasingly to landlords rather than to younger homeowners.

But for all this pain for younger households, high house prices are bringing little benefit to older people since so few are unlocking their housing wealth during their own lifetimes. Instead housing wealth is driving intra-generational inequalities for young and middle-aged adults by increasing the size of bequests. While more people can now expect to inherit wealth – and this represents the only chance many people aged under 45 have to buy a home – the size of bequests is distributed highly unequally. Indeed wealth inequalities are at their highest amongst 25 to 54 year-olds and this is likely to rise with time.

The place of government

The Fabian study demonstrates the critical role of government in securing the wellbeing of mid-income older people. Overall older people ‘in the middle’ consume the equivalent of 99 per cent of their income in social security and public services. Universal entitlements are absolutely essential for this group, making up 70 per cent of their income, when you take account of the NHS as well as the state pension. This underlines the vital importance of universalism as part of the long-term future of the welfare state.

But this is not to say that every universal entitlement should be sacrosanct, as long as reforms do not undermine the wider principle of universalism. The benefits which are today subject to most criticism make only a small contribution to mid-income older people’s incomes (around three per cent, including Winter Fuel Payment, free bus travel and free TV licenses). Removing these entitlements would not be a financial disaster for mid-income older people, although at the same time means-testing them would only save a few billion pounds (and would also restrict their take-up among low income pensioners).

Looking beyond cuts and the question of equality of ‘pain’, age differences in the consumption of social security and public services do not suggest intergenerational unfairness. Mid-income older people receive far more social security than working age households with the same incomes, because of the state pension, but this is something each generation can expect for themselves in time. On the other hand, they consume almost exactly the same level of public services as the average working-age household with the same income. This surprising finding is important for maintaining support for services amongst all age-groups.

By contrast, when it comes to taxation, our project revealed really significant intergenerational unfairness. Retired middle income households pay 27 of their gross income in tax compared to 33 per cent of non-retired households with the same income – and there is a similar pattern across the income distribution. Indeed if retired households paid the same proportion of their gross income in tax as non-retired households government revenue would rise by £7.2 billion each year (equivalent to £1,000 per older household). Moreover, this calculation is based on income alone, before taking account of the greater average wealth older people have, so it should be considered a lower bound for the level of ‘under-taxation’ that currently exists.

Implications for government

The main conclusion of this study is that government should adopt a ‘presumption of equality’ when considering public policy for different age-groups. All policies which appear to give special advantages to older people as a category should be reviewed, because in financial terms alone, older people are no longer special. That does not mean scrapping all age-based variations, but it does mean assessing the evidence and rationale for existing rules on social security, taxation and the design of services.

There may well be times when older people’s higher chances of having significant social or health needs justify age-specific measures, but alternatives to blanket policies should also be considered. And when eligibility is to be determined by age alone, detailed thought should be given to the right age cut-off. Where a policy is designed as a response to relatively low income or high social or health needs it will often be wrong to begin entitlement at state pension age as opposed to, say, 80.

Adopting a presumption of equality is not however an argument for a significant extension of means-testing. This research shows the essential place of the welfare state not just for people facing poverty but for the wellbeing of older people ‘in the middle’. In particular state pension policies make a huge difference to people’s incomes. To guarantee that mid-income older people maintain their current position the government will need to maintain its current approach of a universal state pension uprated in line with rising living standards. The coalition’s current proposals for a flat-rate state pension should also be embraced by all political parties which wish to improve the incomes of future mid and low income pensioners. This measure is cost-neutral with the losses falling on future high income older people.

On the other hand, a presumption of equality calls into question the government’s current ‘triple lock’ policy for the state pension, which guarantees pension increases will be the greater of inflation, average earnings or 2.5 per cent. The Office for Budgetary Responsibility estimates that this policy means that over time the state pension will rise annually by an average of 0.26 per cent more than earnings (and by much more in the short term). With the incomes of retired people converging on those of other households, increasing state pensions by more than earnings on a permanent basis creates significant intergenerational unfairness. The ‘triple lock’ system should therefore be scrapped, although it could be replaced by a smoothing scheme which guarantees against very low pension uprating in certain years, while ensuring that pensions do not rise by more than earnings over the medium term.

In the short-term there is also the uncomfortable question of whether older people should be shouldering a larger share of the burden of deficit reduction. A presumption of equality would suggest that government should avoid any further protection for older people collectively, where this is at the expense of younger age-groups who have already suffered more. That might mean restricting universal benefits like Winter Fuel Payment, Free TV Licences and Free Bus Travel. A perfectly reasonable case can be made for these entitlements when times are good; but the consequences of removing them would be less severe than the impact of many spending cuts which have already been applied to younger households. To save money the age limit for these entitlements could be increased to a point like 80 where middle incomes are lower and typical social needs greater.  Alternatively the entitlements could be means-tested and offered only to those receiving Pension Credit.   There are downsides to either approach in terms of the small but noticeable loss for mid-income older people and the more serious consequences for some low income households. But these are probably outweighed by the principle of equality of ‘pain’.

However the most significant implications of a presumption of equality are found in the domain of taxation. It is very hard to find explanations to justify large age-based differentials in tax, given that the calculation of the tax burden already takes account of older people’s lower incomes. This ‘tax gap’ is even more questionable when we recall the evidence on household wealth. This study therefore recommends that politicians adopt a policy of equalising the burden of taxation paid by people of different ages.

This would need to be a long-term project to avoid a sudden fall in living standards and could probably only to achieve in full as part of comprehensive reform of the tax system.  One option would be to tax wealth and consumption more and earnings less. Alternatively the tax system could be designed to treat earnings and pensionable income the same, by merging national insurance and income tax. These policies are major long-term changes which would need to be designed in a way that avoided harming people on low incomes of all ages.

In the meantime, government should consider some standalone reforms to end the special treatment of older people in the tax system. In 2012 George Osborne announced the end of age-related income tax thresholds for pensioners (a sensible policy which caused a media outcry). Other immediate options would be to apply national insurance to earnings after state pension age and to end tax-free lump sums on private pensions. None of this will be free from controversy, but a gradual, staged increase in the tax burden on older people could be presented as part of a ‘grand bargain’ to pay for universal health and wellbeing services in our rapidly ageing society.

Taxation policy is also critical to the future of the housing market. If house prices do not fall relative to typical earnings then unacceptable intergenerational injustice will become entrenched. So to avoid homeownership falling further amongst younger age-groups policy makers should adopt a target for the affordability of first homes and implement policies in its pursuit. Measures should include increasing the supply of housing and taxing property wealth to suppress rises in asset prices, for example with a Land Value Tax or reformed council tax. These policies would reduce intra and inter-generational inequalities at the same time. However mid-income older people will not be able to pay increased property taxes out-of-pocket. One solution would be to tax transactions – inheritances, lifetime gifts and capital gains on the value of the first home. All these options are politically controversial, however, so an alternative approach might be to replace council tax with a proper annual property tax, but then roll-up the liability of older households into a charge against the eventual sale of their property.

Finally, government should think much more imaginatively about how it can help older people in the middle raise their quality of life by spending more of their money. This should include enabling people to decumulate wealth during retirement and addressing ‘under-consumption’ among the many older people who do not even spend their income in full. The solutions to these challenges could be the topic of a whole project in itself, but there are a number of clear strands for government to pursue. First a serious ministerial initiative is needed in order to massively scale-up the tiny equity release market. An official review should consider the design and distribution of products but also the wider environment, including improving clarity regarding people’s future financial commitments and ways to create institutional structures that are help shift attitudes and behaviour. A similar approach could be considered with respect to the under-spending of income. Of course the balance between spending and saving is mainly a private matter, but it becomes a public policy concern when under-spending contributes to impairment, ill-health, isolation and housing problems.

Implications for service providers

The numerical findings in this report may do little more than confirm the observations of many service providers working daily with older people in all their diversity.  But it’s worth pulling out the implications of two key points: that vast numbers of older people are ‘neither rich nor poor’; and that there should be a ‘presumption of equality’ between age groups. Together these two principles represent a challenge to traditional models of service delivery which once assumed their users would be both old and poor. Of course times have changed and there are excellent services which assume nothing of the sort, but elements of the stereotype persist.

The ‘presumption of equality’ means first that service providers should seek to confront any subconscious ageism or infantilisation in their assumptions about the older people they work with – as well as seeking to expand their clients’ sense of possibility for their own lives. It also means that services must ask themselves whether they should be geared exclusively to a specific age group, or whether other means of identifying whether a service will be suitable and beneficial are more appropriate. In particular, providers should question whether the state pension age is ever a suitable cut-off for offering services, since needs change incrementally and reaching state pension age is not linked to a sudden drop in income.

For housing providers the debate about age-exclusive services partly relates to equitably addressing social need; small well-designed flats close to local amenities might appeal to a wide range of people after all. But it also touches on the nature of the communities in which people wish to live. In this study we have not looked at the evidence for and against age-exclusive communities, but at the very least it would seem desirable to ensure that people needing specialist housing have a choice including the option of living in communities with a broad age mix.

Turning to incomes, service providers need to work on the basis that many of their service users will have a reasonable disposable income and/or other significant assets. They should avoid a ‘means-test’ mentality where people ‘in the middle’ are encouraged to fend for themselves, because the pressure is already high from the poorest older people.  This phenomenon is well documented with regard to social care, with councils avoiding their duty to assess and offer support to all who need help in their homes. There are anecdotes of this with respect to accessing retirement housing in the social sector too, although our analysis of the ELSA data did not show a huge difference between the number of low and mid income households living in rented retirement housing.

Where there is pressure on resources, instead of covert rationing service, providers should expand the use of customer charges and fees. The challenge is to design services that work for everyone, with hybrid funding models bringing together personal payment and external sources of funding, including different funding arrangements for people with different resources using the same services. This approach will minimise the chance of two-tier provision developing, with means-tested services for the poor; premium paid services for the rich; and middle income groups falling between two stools.

Service providers should also feel it is legitimate to work with older people in the middle to increase what they spend. Providers could offer support to explain the services and activities that can be bought and help people make choices. For example they could offer specific support to help people maximise the value of their disability benefits, in a light-touch version of the support social care clients receive alongside a personal budget. When people have insufficient income, providers could also do more to help them work-through their options for unlocking capital including downsizing and equity release. People can be very emotional about preserving housing wealth but providers should think of ways to encourage people to make the most of their assets on an ongoing basis, rather than seeing them as only a fund for an emergency. One possibility might be for financial companies and housing services to design hybrid products that unlocks capital and uses it to directly fund services or support.

People are resistant to unlocking housing capital partly because of the housing alternatives on offer. Better options for ‘downsizing’ need to be developed that both unlock a decent quantity of money and provide an attractive home that people actively want to move to. For some this will mean retirement housing, but demand from mid-income older people is constrained by tenure and affordability. People rarely change tenure when they move and most retirement housing is rented while most mid-income older people are homeowners. People are also understandably worried about the charges associated with retirement housing, which can lead to a 20 per cent fall in disposable income for someone with a middle income. Housing providers should consider developing new models which work for homeowners with relatively modest incomes but who would not be unlikely to be eligible for housing benefit. For example, providers develop mixed-tenure communities, with the offer of an option where monthly charges are rolled-over against the eventual sale of the home.

This study has focused more on diagnosis than solutions and these are tentative suggestions on the direction for service providers. But what is completely clear is that change will come, from a new public policy environment and from the evolving complexion of the older population. Service providers should embrace the rise of ageing in the middle and the new presumption of intergenerational equality.