Determining Life Expectancy: Not Just for Actuaries

Determining life expectancy is an important calculation, and not just for actuaries and life insurers—pension professionals and plan sponsors have a stake in it too. A recent webinar discussed a variety of considerations concerning life expectancy. 

Determining life expectancy is “important because there’s such a wide range of life expectancy for different pension plans across the United States,” remarked Dan Reddy, CEO of Club Vista US, an independent data utility that supports pension funds, advisors, insurers and asset managers. He was joined as a panelist in the National Institute on Retirement Security (NIRS) webinar “Mortality Improvements for U.S. Defined Benefit Pension Plan Participants” by Club Vista Chief Content Officer Erik Pickett and NIRS Executive Director Dan Doonan. 

Calculating Life Expectancy

Reddy identified two steps in calculating life expectancy.   

 

Characteristic	Baseline	Future Trends
Vantage Point	Provides a snapshot of current state of longevity	Looks at how life expectancy will change in the future
Objective/Subjective Measure	Objective	Subjective Measure
Direction it Looks	Past experience	Recent experience is a good starting point, but considers how and when will it change

 

There are additional approaches Reddy cited, as well: 

• Top Down: Collect large amounts of data from similar pension plans; average out experience and apply to plan
• Bottom Up: Calculate longevity for different individuals based on their characteristics; combine assumptions for the individuals within the plan

Factors to Consider

There are a variety of factors that affect life expectancy, Reddy suggested. He said that more than genetics are involved—in fact, he said, genetics play only a 15%-20% role in how long one is expected to live. He said that the following factors are “much more predictive than any genetic information that we have”:

• lifestyle: for instance, using the zip code where an individual lives as an indicator
• affluence: having a pension
• occupation: what type of “collar” describes one’s job
• health: considers whether one retired on disability or under more normal circumstances
• gender

But even these factors are not set in stone, Reddy indicated. He said that Club Vista is “constantly evolving our model to help pension plans understand more and more information about their participants.” For instance, he said, adding salary information to help gauge affluence may be better than just considering a pension amount. In addition, he noted, they found that difference between the there is a 1.5-year difference between the life expectancy measurement based on gender and a 3.2-year difference based on the zip code in which people live.

Reddy indicated that Club Vista has found that it is more useful to model all predictors simultaneously. He further argued that getting a more accurate view of the longevity characteristics of one’s plans by zeroing in on participants can be useful. “Off-the-shelf alternatives do not catch the wide range of life expectancy valuation that we see in the United States,” he said. 

Reddy also suggested factors to consider concerning life expectancy when the time comes that information from the period affected by the pandemic is added to the analysis, including: 

• Were your pensioners geographically concentrated in hard-hit areas? 
• Is your plan more male or female dominated? 
• What is the age profile of your pensioners? 
• Is your plan in an industry on the front lines of the pandemic? 
• Do you offer retiree health care?

Changes in Mortality Rates

Pickett noted that their research has found that having a pension plan can contribute to improved mortality rates. He noted that during the period 2014-2017, annual improvements in standardized mortality rates were higher than those for the population at large, and for both genders. 

 

Group	Improvement in Mortality, U.S. Population as a Whole	Improvement in Mortality, Pension Plan Participants
Male	0.6%	1.4%
Female	0.5%	1.3%

Reddy noted, however, that the Club Vista data does not include information from the period affected by COVID-19. 

Implications for Pension Plans

Increasing longevity “could have a significant impact on pension plans if it is to continue out into the future,” said Pickett. For one thing, it could spell 2%-8% more in pension plan liabilities. “It could have a significant effect on securing pension benefits,” he noted.