Gloom and doom stories are easy to write concerning future ratios of retired population to workers (and I have written such articles). I am not backing off from using these ratios to criticise governments about better balancing the promises of pensions and health care with future revenue capacity. However there is a more positive view we can put to individuals and their advisers who do face up to their own balancing of life expectations and future income.
Over 65 isn't what it used to be
Let’s reflect on this Dependency Ratio measure. It’s generally been calculated as population aged 65 and over divided by population aged between 15 and 64. Alternatively it can be inverted and shown as ‘workers’ (population aged 15 to 64) divided by ‘retirees’ (population aged 65 and over) and that’s the way I will express it here.
When the Australian age pension was introduced in 1909 the ratio was 15.0 (i.e. 15 workers per 1 retiree). If we use the same age brackets and current population data (2011 census data) we get 4.9 workers per 1 retiree. Projecting this population and future expected longevity trends the ratio declines further to 3.5 in 2025 and 2.6 in 2050.
But what if we defined the age brackets for this ratio more dynamically? In 1909 age 65 was probably a fair average for when bodies broke down and mental and physical function made sustainable employment impractical. There are a number of ways of dynamically standardising what this age should be and allow for improvement over time in lifetimes and management of disabilities.
One approach is to calculate Disability Adjusted Life Expectancy (DALE). Many readers would already be familiar with ‘life expectancy’ – average expected future years of life calculated from life tables. DALE is a more complex version of this calculation which measures the equivalent number of years of life expected to be lived in full health i.e. healthy life expectancy. DALE requires quite a lot of data on burden of disease by age and is an evolving methodology. For this reason it’s difficult to go back in time and work out what DALE would have been in 1909.
An alternative to DALE is to standardise the worker retiree ratio for different points in time based on assuming a constant percentage of total lifetime which is spent in retirement. Based on my research of mortality tables, I think this is a good proxy for DALE methodology.
For example, in 1909 when 65 was first used as the retirement age, life expectancy from that age was to live to age 77 (average for men and women). This means the average period expected to be spent in retirement (12 years) was 16 % of total lifetime (77 years).
Moving more than 60 years forward to 1975 life tables, there was not a great deal of change in this position. Using 16% of lifetime expected to be lived in retirement as the standard, we get age 68 as the standardised retirement age in 1975 – just 3 years more than in 1909.
However from 1975 onwards to 2011, the longevity improves at a much greater rate. The standardised dynamic retirement age increases a further 6 years to age 73. This is now a long way from the age 65 we normally use in our statistics.
Now if we use this more dynamic way of calculating retirement age, dependency ratios have a much more stable pattern. From 1909 to 2011 instead of reducing from 15.0 to 4.9 workers per retiree, the ratio stabilises around 9.0 from 1950 until the present time. Allowing for longevity to keep improving in future at the rate it has been recently, we do see some further decline in the ratio to 2025 (7.2) and by 2050 (5.3) but well above the catastrophic 2.5 using unadjusted age 65 as the retirement age basis for the ratio.
What conclusions can we draw?
So what are the conclusions from all this? The following are a few thoughts:
- retirement planning advisers need to consider both total longevity and healthy life expectations
- whilst people (and their employers!) may tire of a ‘major’ career between say age 55 and 65, it doesn’t mean retirement starts and income generation stops when this inflection point is reached. Different, more flexible occupations sustaining basic living costs will need to be planned for by individuals and their advisers until the point when genuine physical incapacity for work arrives
- real value of capital accumulated needs to be protected past the traditional retirement age using endowment fund type strategies, as there may be many years of a healthy lifestyle to come
- longevity insurance products will offer the best value to clients if the income payout is targeted from the ‘dynamic’ retirement ages (using the above methodology, closer to age 73 than 65) rather than starting payments from when clients cease their ‘major’ full time careers. This is the ‘sweet spot’ of deferred annuities (even sweeter now if the government legislates recent announcements)
- if a genuine market now develops for deferred annuity products, published league tables of deferred annuity rates will help educate clients and their advisers much more simply about dynamic retirement age concepts than studying complex mortality tables and talking to actuaries.
It might even help to stop thinking of someone who is 72 as a ‘dependent’.
Bruce Gregor is an actuary and demographic researcher at Financial Demographics and established the website www.findem.com.au.