A conservative rationale for raising the age of eligibility of New Zealand’s public pension.
New Zealand’s Prime Minister announced plans in March 2017 to increase the age of eligibility for New Zealand Superannuation to age 67, providing his government is returned after a pending general election. See more here.
One critical part of the justification for raising the eligibility age is increasing longevity. The Cabinet Paper said…
“Life expectancy has increased by 12 years in the last 60 years, including four years since 2001 when the retirement age became 65. Life expectancy is expected to continue to rise on average 1.3 years per decade until the end of the century – the equivalent of increasing by almost one day each week.”
…and gave the following table:
| Life expectancy | ||||||
| 1951 | 1956 | 2001 | 2017 | 2037 | 2057 | |
| At birth… | 69.3 | 70.5 | 78.7 | 82.5 | 86.2 | 89.0 |
| At age 65… | 13.8 | 14.1 | 18.4 | 20.9 | 23.4 | 25.5 |
The paper did not say whether this is period or cohort life expectancy , but from the data it’s clear it’s period. This means that, for example, a 65-year-old living in New Zealand in 2037 would be expected to live a further 23.4 years on average, assuming no further change in mortality rates after that date. However, it does take into account expected changes in mortality rates up to 2037.
As covered in blogs here and here, period life expectancy is the most widely used measure of the average health and death rates in a population. However, cohort life expectancy is a more intuitive estimate of the expected lifespan of the individual representing the average in a population.
A rationale using estimated future lifespans
To understand the expected future lifespan of the average 65-year-old New Zealand resident in 2037, we need the cohort life expectancy of that individual. This cohort life expectancy assumes changes in death rates up to 2037, and in the years after 2037.
Such assumptions exist. They are necessary for the population projections produced by Statistics New Zealand, who also publish cohort life expectancies. Given that the paragraph above the table alludes to life expectancy continuing to increase until the end of the century, which is an outcome of assuming that death rates continue to decrease over that timeframe, it would be consistent to look at cohort life expectancies.
Here are the cohort life expectancies at age 65 for New Zealanders who reach age 65 in each year of the table in the Cabinet paper:
| Cohort life expectancy at age 65… | ||||||
| …the average expected remaining lifespan for those reaching age 65 in the year: | ||||||
| 1951 | 1956 | 2001 | 2017 | 2037 | 2057 | |
| Assuming medium death rates | ||||||
| Female | 15.9 | 16.3 | 21.5 | 23.8 | 26.0 | 27.7 |
| Male | 12.9 | 12.7 | 18.9 | 21.3 | 23.9 | 25.9 |
| Assuming low death rates | ||||||
| Female | 15.9 | 16.3 | 21.6 | 24.4 | 27.2 | 29.4 |
| Male | 12.9 | 12.7 | 19.1 | 22 | 25.1 | 27.7 |
Source: Statistics New Zealand, How long will I live?
The table looks more complicated, even though it only gives cohort life expectancies at age 65. This is partly because I have given male and female life expectancies separately as is the near-universal practice. The other difference is the need to state the assumptions on the trajectory of future death rates. The “medium” assumption is the central estimate from Statistics New Zealand who also give a “low” assumption which is more optimistic on how fast death rates will fall in future.
The differences between period and cohort expectancies are zero or small for the oldest cohorts. The differences are greatest for the youngest cohorts as they have more future life in which the assumption of reducing death rates takes effect. The cohort life expectancies are higher than period life expectancies because the assumption of continuing decreasing death rates mean a lower chance of dying at each age which leads to longer expected lifespans.
Does it matter?
If cohort life expectancies using assumptions of medium or low death rates were used instead of period life expectancies, then people aged 65 in 2017 or later would see an estimate of their expected lifespan after age 65 of around 1½ to 3 years longer.
The Cabinet paper suggested that a retiree at age 65 can currently expect on average to live 24.3% of their life on NZ Super and the first people with an eligibility age of 67 years would have slightly longer at 24.5%. Using cohort life expectancy would put the current estimate nearer 26% and again slightly higher for the first people to have 67 as an eligibility age. (The precise figures are: current 25.8% on medium death rates and 26.3% on low death rates; for the later cohort 25.9% to 26.9%.)
The figures for life expectancy used in the Cabinet paper underestimate how long people might live compared to what are the closest-to-official projections for how longevity might change in future in New Zealand. Does it matter? The differences are small, but significant in two ways:
First, in being conservative, the government cannot be accused of over-reacting to longevity trends. The timetable to increase age of eligibility is benign compared to other countries.
Second, an opportunity is missed to tell the story on how lifespans are expected to increase. The media coverage on the planned age of eligibility increase was a chance to inform people about lifespan prospects. Underestimating how long people might live increases the risk that retirement planning is unrealistic.
*****See an update on New Zealand’s policy on the age of eligibility here*****
The cohort life expectancies in the table for this post were downloaded from How long do I live? on 11 April 2017 and are based on Statistics New Zealand, complete cohort life tables 1876–2015 (updated March 2016) and national population projections 2016(base)–2068, mortality assumptions (published October 2016).