Period Life Expectancy vs Cohort Life Expectancy: The Difference is Important

The Encyclopaedia Britannica website explains that life expectancy is: “an estimate of the average number of additional years that a person of a given age can expect to live.” The key word to focus on here is average, and there are different ways of calculating it.  

How is life expectancy calculated?  

There are two main calculation methods used: period life expectancy and cohort life expectancy.  While these may sound technical, it is important for people working in financial planning, superannuation, the media, etc., to understand and appreciate the difference. 

Period life expectancy 

Period life expectancy is the simpler of the two methods, and it’s the one typically quoted in the media and in legislation. However, it can potentially provide a significant underestimate. Part of the reason it’s so commonly used is it’s easier to look up and the life expectancies don’t change until the next life tables are produced.  

Period life expectancy figures are just one part of the latest Australian Life Tables published by the Australian Government Actuary (AGA) every five years. The most recent life tables are based on the three calendar years centred on the 2016 census (the ‘period’).  

The AGA looks at the total number of deaths in those three years by age as a percentage of the number of people who were alive at each age. This data informs a set of mortality rates for the Australian population and these are what life expectancy calculations are based on.  

Mortality rates give us the probability that a male or a female of any age passes away in the year before their next birthday. The probability an average 30-year-old passes away within a year is under 0.1%. But for older people, the probabilities increase markedly. The probability an average 90-year-old man dies within a year is around 15%. 

By looking at the mortality rate at each and every future age of a person’s life, we’re able to calculate the probability they will survive to each future age. Taking a 60-year-old, for example, we calculate the probability they will survive to age 61, then to age 62, then age 63 and so on, all the way up to the end of the Australian Life Tables – age 109. This lets us calculate the average age they’ll live to – based on all these probabilities.  

For a 60-year-old male, using the latest Australian Life Tables, his period life expectancy is to age 84.0; for a 60-year-old female, it is 86.9.

It is called their period life expectancy because the mortality rates used came from a specific period. In this case, from calendar years 2015 to 2017. 

Cohort life expectancy 

Because of ongoing medical developments and trends toward more and more healthy lifestyles (e.g., reduced levels of smoking), mortality rates have been decreasing steadily over past decades, particularly since the 1970s. It means that life expectancy calculations based only on mortality rates from a specific period in time are plain wrong for most practical purposes.

Actuaries don’t expect that the number of years a person will live is their period life expectancy. Instead, we also need to consider these improvement trends. This is where cohort life expectancy comes in. 

Cohort life expectancy allows for the fact that mortality rates continuously change during the lifespan of a group of people – or cohort—as they get older. Improvements in medical treatments and healthier lifestyles generally result in a reduction in their mortality rates from one period to the next.  

Each time the Australian Life Tables are updated, the AGA carefully analyses how the mortality rate at each age has been changing over time by looking at data for up to the last 125 years.  

More recently, each set of Australian Life Tables comes with accompanying estimates of how mortality rates may change (improve) into the future. These assumptions are called mortality improvement rate assumptions. 

Cohort life expectancy calculations allow for this extra information, resulting in a significant improvement in accuracy.  Instead of looking up static mortality rates from, say 2015, which effectively assumes each person spends their entire life in 2015, we acknowledge that people reach each birthday in a different future calendar year, with mortality rates that are expected to have changed (reduced) relative to when the tables were produced.  

Chart 1 shows an example of how mortality rates are projected to change over time. It shows the estimated mortality rate at age 75 over time¹.  

For a 65-year-old in 2024, by the time they reach age 75 (in 2034), the estimated mortality rate for 75-year-old males would be nearly 40% lower than when the tables were published 18 years ago.  

Chart 1: Reducing mortality rate for a 75-year-old male 

In Chart 2 below, each column shows the period life expectancy for a 65-year-old male using the mortality rates from historic life tables. You can see the strong, steady improvement upwards in the blue bars, reflecting the difference in mortality rates observed between each period. The red line shows their cohort life expectancies – which are higher because the person’s assumed mortality rates will continue to reduce in future. 

Chart 2: Period and cohort life expectancy of a 65-year-old male

Using the latest Australian Life Tables with the 25-year improvement rate assumptions, the cohort life expectancy of a 60–year–old male is age 87.9—about four years more than his period life expectancy. For a 60-year-old female it’s age 89.8—about three years more than her period life expectancy. 

Estimates are only that, however. Actual rates of improvement can vary from the estimates. Variations can occur if the impact of positive future medical advances are different to the estimates or if negative changes occur, such as increases in obesity, drug deaths, pandemics, etc.

Concluding comments

Life expectancy is not a specific prediction of how long any particular individual will live—it’s just an average for a group or cohort at each age.

None of us are average. We are all different with different lifestyles, different genes, different attitudes to risk etc. So, the average is only relevant to a very small percentage of people.

Cohort life expectancies are expected to provide a more realistic representation of the average number of years an individual or a group of people will live. However, as you can see, they can be more complex to calculate and not as easy to simply look up in a table—as they depend on the year that the calculations are done.

Period life expectancies can be easier to look up but, they are based only on mortality rates that applied during a certain historic period without any further allowance for future mortality improvements which can be significant.

Don’t forget: Life expectancy calculations are based on the range of lifespans that can transpire for the individuals within a group. They are not a prediction. There can be a large standard deviation around average life expectancy figures. For new retirees, the standard deviation is roughly eight years.

For financial planning, it’s vital to recognise that some people will have shorter lifespans, and others will have very long lifespans. What happens to each individual can be subject to randomness. A more appropriate rule of thumb might be to focus on the age when most people in a group will have passed away. For more information, read the Actuaries Institute’s Research Note, The Importance of Accurate Life Expectancy Calculations in Retirement Advice.

For a calculator that takes all these considerations into account, please see the Optimum Pensions lifespan calculator or the Canadian Financial Planning standards. 

An extra insight: The impact of COVID-19 on mortality trends

We anticipate that the next Australian Life Tables, published around the end of 2024, will provide an analysis of the impact of COVID on life expectancies. Early indications are that the impact may be less than some commentators anticipate.

Table 1: Total deaths in Australia aged 60 or more

Chart 3: Total deaths over age 60 in Australia

References

CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.