Retirement savings – how much is enough?

How much is enough?

NEST Insight is looking for solutions to a central challenge when it comes to retirement savings: how much is enough? If you have an answer – or even part of an answer – to this challenge, we’re keen to hear from you.

The challenge of adequacy

Thanks to auto enrolment, millions of people are saving for retirement, often for the first time. By the spring of 2019 those who’ve stayed enrolled will be putting aside at least 8% of a band of earnings into a workplace pension. This is a major step towards fixing the shortfall in retirement savings. Yet many argue that people should be saving more than this 8% minimum – perhaps significantly more.

One of the central challenges facing the pensions industry is the question of adequacy. How much should members of defined contribution (DC) plans put aside for their retirement? This question is currently top of mind for policy makers, pension trustees – and anyone with an interest in getting good outcomes for DC savers.

The challenge is even greater when these savers have been enrolled automatically. Their engagement is inevitably lower than those who’ve chosen to enrol. This can leave trustees and policy makers having to make decisions on their behalf. For instance they might choose to increase the default or minimum rates at which savers contribute. This can encourage some people to save enough but it can also mean that other groups end up oversaving. In the absence of detailed knowledge of each individual saver’s personal circumstances, how can those responsible for pension schemes make effective decisions on their behalf?

In search of solutions

NEST Insight is looking for practical solutions to this complicated challenge. In June we held a session to discuss and debate potential approaches as part of our inaugural conference. We’re grateful to panel chair Rowena Mason (Institute from Fiscal Studies), and to contributors Cesira Urzi Brancati (International Longevity Centre), Charles Woodhouse (Q-Super) and Bonnie-Jeanne MacDonald (Dalhousie University) for their contributions.

The session was a valuable way to tee up the issues and consider alternative solutions. We’re now keen to hear views and proposal from interested parties in finding practical solutions that can be delivered in the real-life context of a workplace pension.

The first major challenge is that people’s consumption patterns in retirement are complex and diverse – and they change over time. This can make it hard to set clear target income rates for savers. In the panel, Cesira Urzi Brancati noted that consumption falls with age. A household headed by an 80 year old spends half as much as a household headed by a 50 year old, and consumption falls most sharply after the age of 65. However, it’s not clear to what extent this is caused by declining assets, or by changes such as reduced levels of activity. Furthermore, the overall trends disguise significant variations. Older people are not a homogenous group. Differences are important, and are driven by taste as well as income. With this in mind, Cesira has developed a set of segments reflecting different patterns of consumption.

The challenge is even greater when these savers have been enrolled automatically. Their engagement is inevitably lower than those who’ve chosen to enrol. This can leave trustees and policy makers having to make decisions on their behalf. For instance they might choose to increase the default or minimum rates at which savers contribute. This can encourage some people to save enough but it can also mean that other groups end up oversaving. In the absence of detailed knowledge of each individual saver’s personal circumstances, how can those responsible for pension schemes make effective decisions on their behalf?

Even if future consumption needs were clear, there are significant challenges around the information available during the years when people are still saving for retirement. On the one hand, pension schemes have imperfect data. They tend to know a lot about some aspects of their members’ lives – for instance age and current level of contributions – but very little about their wider financial or household circumstances. This makes it hard for schemes to make decisions on their members’ behalf.

The parallel challenge is a lack of information and capability among pension savers. This makes it hard for them to make decisions for themselves. In chairing the panel session, Rowena Mason set out the difficulties facing individuals trying to calculate what they will need in retirement. She suggested that there are two ways in which the industry can support individuals:

  • by providing information and tools to support decision-making
  • by making calculations on their behalf.

Both methods are fraught with challenges – and they may not be mutually exclusive, as our panel discussed at the session.

Generic versus tailored approaches

Whether individuals make their own decisions, or scheme do so on their behalf, there remains the challenge of calculating a suitable target income for retirement. This could be done by setting a single benchmark for a whole population. One common example is the ‘replacement rate’, which defines a percentage of a saver’s pre-retirement earnings that they should aim to receive as pension income. Inevitably, though, this kind of blanket approach fails to take into account the needs of different people living in different circumstances when they retire.

Alternatively, more tailored solutions can be developed, using algorithms to calculate personalised targets based on a wider set of data about the saver’s circumstances. These approaches, though, require engagement from participants, and a willingness to provide data on their household finances.

In Australia, where the compulsory superannuation system has been in place for some years, a 70% replacement rate has become a standard benchmark. In our panel, though, Charles Woodhouse explained that this can be a difficult measure to translate for individual Q-Super members, especially those who work shifts or part time patterns. Q-Super’s goal is to get members into personalised plans, both while they save and while they draw down income in retirement. They aim to move away from one-size-fits all approaches, using the information they already hold about their members; but this depends on members engaging.

Bonnie-Jeanne MacDonald took this challenge one step further by contesting the 70% replacement rate. She said that no-one has adequately tested whether such a ratio leads to good outcomes. The problem is not the 70% figure itself, it’s that income in a single year does not reflect an individual’s standard of living. Standard of living is determined by many other inputs such as household make-up and needs, partner income and levels of debt. Bonnie-Jeanne therefore suggests a new measure: Living Standards Replacement Rate (LSRR), where the aim is for people to have as much to spend on their lifestyles as they did before they retired. We believe this Living Standards Replacement Rate deserves further investigation. For more information on this approach, please see this interview with Bonnie-Jeanne.

Next steps

Despite the importance of securing an adequate retirement income, it seems there are currently no clear and proven solutions to working out ‘how much is enough’. There is a trade-off between what industry providers would ideally like to review for every individual and what they can pragmatically consider. And of course, there is no fixed age where retirement starts and consumption changes.

‘Adequacy’ is a key research theme for NEST Insight, and our programme over the coming months will seek to explore suggestions from this panel as well as from the wider industry and policy-making world. We welcome your contribution to this debate by email or in the comments below.