Dealing With Age Inflation

My op-ed today proposes a systematic fix to the lobbyist-laden bunfight that seems to accompany every proposal to change a statutory age limit. I argue that we should index legislated ages to longevity improvements. Full text below.

You’re Only as Old as they Feel, Australian Financial Review, 6 October 2009

On his 64th birthday, Paul McCartney suggested that he might change the lyrics of one of his most famous songs to “When I’m 84”. While most took this as a sign that old age is in the eye of the beholder, it also reflects the fact that 64 isn’t 64 anymore.

The Australian life tables tell us that a 64 year old in the Beatles Era could expect to live for another 15 years. Today, a 64 year old can expect to live for 21 more years – an improvement of 6 years. According to work by Harvard’s David Cutler and co-authors, the improvement in health has been even greater. They find that the self-reported health of men aged in their early-70s today is similar to the levels recorded for men aged in their early-60s three decades ago.

So in terms of life expectancy, 70 is the new 64. And in terms of health, 74 is the new 64.

With a steady decline in the share of backbreaking jobs, and ongoing improvements in medical technology, it is reasonable to expect these numbers to keep rising. Yet to look at many of the statutes on our books, you would think that none of these changes had ever occurred.

The one area in which government appears to have correctly accounted for rising longevity is the eligibility age for the pension, which is now scheduled to rise to 67 in 2017. Yet the federal government has poured cold water on suggestions that the superannuation preservation age should also rise: a decision that is both inefficient (given longevity increases) and inequitable (since it advantages richer superannuants over poorer pensioners).

When each proposal to raise age limits is taken in isolation, it is no surprise that policymaking will be ad hoc. A better approach would be to index upper age limits in all laws, ensuring that they rise together as lifespans increase.

We already have a precedent for this kind of across-the-board indexation: the use of ‘penalty units’. Rather than write specific dollar figures into legislation, Australian parliaments typically set maximum fines as a certain number of penalty units. The intuition for this is straightforward: a fine of $1000 in 2009 is considerably less harsh than a fine of $1000 in 1999. Updating penalty units to keep pace with inflation is a straightforward way of ensuring that fines remain constant in real terms over time. It also helps ensure that the relative magnitude of different fines (eg. tax fraud and welfare fraud) remains unchanged.

How might age indexation operate in practice? One approach would be to mandate that all elderly age limits should increase by 3 months every year (approximately the rate at which life expectancy is presently rising).

More radically, we could define ages in terms of time from death rather than time from birth. For example, we might legislate so that the typical person is eligible for the pension for the last 20 years of their life (which on current life tables would be age 65½). While this approach is a tad morbid, it does have the advantage of focusing directly on the policy parameter of interest, namely the expected number of years that a law will affect the typical individual.

Age indexation should apply not only to laws that provide special benefits (age pensions, Veterans’ pensions, superannuation), but also to legislation that imposes additional requirements on the elderly. For example, drivers in many states are presently required to undergo annual tests after a certain age. As longevity improves, this age should steadily be shifted upwards. Similarly, laws that allow exemptions for the elderly (such as jury duty or voting) should have their age limits indexed so that they steadily rise over time.

Naturally, such indexation would only apply to upper age limits, since rising lifespans do not strengthen the case for lowering the age at which young people can leave school, drink or vote. If anything, longer lifespans suggest that such ages should be raised rather than lowered (today’s 18 olds will vote in 3 more federal elections than when the voting age was first lowered to 18 in 1973).

With limited political resources, parliament should aim wherever possible to find across-the-board solutions that are fair over time and across groups. Indexing fines and allowances has allowed politicians to spend less time debating dollar figures, and more time discussing the big issues of the future. Likewise, a move to age indexation could avoid the unseemly battle that arises each time a particular age change is suggested. Rather than age limits being determined by lobbyist power, wouldn’t age indexation be a fairer approach?

Andrew Leigh is an economist in the Research School of Social Sciences at the Australian National University.

9 thoughts on “Dealing With Age Inflation”

  1. I doubt this will work since it is age at which health frailty sets in not age of death that determines ability to work. People are living longer for sure but the age at which they become frail is not increasing nearly as much.


  2. Hc, I think andrew’s point in “They find that the self-reported health of men aged in their early-70s today is similar to the levels recorded for men aged in their early-60s three decades ago.” is that at a given age the health of the average person is improving compared to previous generations.
    This is a good principle you are proposing Andrew.  It prevents all sorts of “creep” due to changing fundamentals.  It also somewhat compensates for changes like the average age of starting full-time work or average age of getting married or having children.  All of these have drifted slightly later in life.
    It really should be applied to pretty much all dollar amounts with regards to inflation (or possibly change in average disposable income) e.g. Land tax or stamp duty thresholds.


  3. I don’t know about ‘self-reported heath status’ but my understanding is that expectation of life at age 65 has barely shifted in recent years and that at age 85 it has not shifted at all.

    Isn’t that the best way of assessing health at advanced ages.


  4. hc, we’re in furious agreement on the right measure (see my 2nd para). That measure has not only risen since the 1960s, it’s still rising (eg. by 2 years over the period 1997-2007). Just click on the “Australian Life Tables” link above for full data if you’d like to play around with it. Year-to-year movements are a bit noisy, but over a 5 or 10 year period, the patterns are pretty clear.


  5. It’s closer to four months a year than three months a year.  If we could just do three times better, we’d all live forever. And yes, all age limits should be floated. Also all monetary figures – fines, taxes, etc. Anything else is a waste of good computing power.
    It’s also worth noting that the fact that we are going up by four months a year – spread out over all ages – does seem to demonstrate that the ‘obesity epidemic’ is not in fact doing us that much harm.  It simply doesn’t show up in the graph.


  6. Andrew, I can’t access these tables as don’t have password  but these

    I can. They show life expectancy of white US males 1850 to 2004. At birth lifespan almost doubled from 38.3 years to 75.7 years.  At age 60 it went from 15.6 to 20.9 years and at age 80 from 5.9 to 8.1 years.

    Your data might be better and ‘morbidity compression’ may have set in but in the very long term at least we are dramatically increasing lifespans but not greatly increasing the lifespan of those who the old. 

    In short (on the basis of these figures) we are better at getting people to live longer but less able to get them to live longer, healthier lives.  Or am I seeing things wrongly?

    It could be that both sets of data are correct. Morbidity at older ages seems seems to have improved most emphatically since the 1960s. From 1850 to 1960 it barely moved.


  7. Morbidity at older ages seems seems to have improved most emphatically since the 1960s. From 1850 to 1960 it barely moved.
    This is probably because the big initial improvements in life expectancy were due to big reductions in infant and childhood mortality.


  8. 1. I think these days we can get much more accurate about when someone could be expected to die. Everyone should be able to expect 20 years of retirement. Smokers, people with genetic diseases and extreme sports-people should be able to retire much earlier.

    b. Should some of the benefit of advances in health actually be apportioned to the people? Why shouldn’t we have longer retirements? We’re still working the same hours each day despite there being little real work to do because of industrial machines.


Comments are closed.