Defining what makes an ‘adequate’ retirement income is always going to be tricky. It’s inherently difficult to know exactly what people’s spending choices and needs are likely to be, or how they will adjust to stopping work.
Add in the changing nature of retirement, where increasing numbers of people are working past their State Pension age, it becomes even harder.
New paper, new ideas?
The Institute for Fiscal Studies (IFS) recently published a paper investigating a new method of looking at retirement incomes. It identifies an ‘optimum’ level of pension saving for each couple household. Instead of income band, this is based on a range of personal circumstances (e.g. number of children) and an
assessment of spending patterns. It then evaluates whether people have saved below, at, above or the ‘optimal’ level required to achieve a comparable standard of living for their retirement.
This is an innovative approach, and one that Age UK welcomes as an important addition to the retirement income adequacy debate. However, as with most modelling, there are potential flaws. The IFS does acknowledge many of these, for example the assumption that people judge an optimal income based on purchasing power rather than ‘keeping up with the Jones’’ or don’t factor in leaving bequests.
And the sample used is potentially misleading when trying to form general headlines – it is based on a particular cohort of people born in the 1940s, so is unlikely to reflect anything other than a snapshot.
Furthermore, the conclusions presented in the initial paper are very different to the existing evidence about retirement saving.
It finds that 92 per cent of the cohort examined (couples born in the 1940s) have ‘over-saved’ for their retirement, based on comparing actual savings with the model’s estimate of the optimal level for each individual, while 42 per cent would optimally hold no private pension wealth.
It just doesn’t match other evidence. For example, the Pensions Policy Institute calculated that among older workers (aged 50-64 in 2011 – a similar cohort to that in the IFS research), only 40 per cent would achieve an ‘adequate’ retirement income by their State Pension age.[i] This simply does not tally with the IFS’ findings.
We are concerned that this analysis seriously underestimates the amount of money needed in retirement. To their credit, the authors are transparent about their method, and identify some important gaps in their own model.
For example, spending on long-term care is not accounted for (as the authors acknowledge). With a median average spending of £31,700 per person[ii], it is clear that this is a major cost faced by many people in retirement and would make a huge difference.
There are also different ways of approaching the issues of calculating average incomes. This model uses the average income between the ages of 20 and 50 as the benchmark, ignoring everything from 50+. Typically, income closer to retirement, for example between 50 and 64 is accounted for. In addition, it’s based on a cohort born in the 1940s, it is an estimate that relates only to this group of people, and is not future-proofed.
We agree that there are problems with using set benchmarks and that other ways should be explored. And despite of some of the discrepancies in its findings, the IFS’ study could potentially represent a really important innovation in measuring retirement income adequacy. But in itself it is not a sound basis for policy-making.
[i] Pensions Policy Institute (2012), Retirement income and assets: the implications for retirement income of Government policies to extend working lives
[ii] Comas-Herrera A & Wittenberg R(2010), Expected lifetime costs of social care for people aged 65 and over in England, PSSRU