Uncertainty about mortality will rattle life, health and pension sectors
The life expectancy of Americans used to go up almost every year. Now, it doesn't. And it's not just because of COVID-19.
The trend
Life insurers will have to prepare for the possibility that the life span of life insurance policyholders, annuity holders and pension plan participants could fall.
The driver
U.S. life expectancy at birth increased, or at least stayed the same, almost every year from 1943 through 2014.
Insurers, actuaries and personal financial advisors have acted on the assumption that the key risk most family breadwinners and retirement savers face is longevity risk — the risk that retirees will live longer than expected and use up their assets.
But between 2014 and 2015, something changed: Possibly because of the opioid epidemic, overall U.S. life expectancy at birth fell to 78.7 years, from the 2014 peak of 78.9 years.
Life expectancy at birth stayed the same or fell every year for the next three years.
The National Vital Statistics System, the arm of the National Center for Health Statistics that compiles U.S. birth, death and life expectancy statistics, changed its methods for calculating life expectancy in 2018. Life expectancy then crept up to 78.7 years in 2018, and to 78.8 years in 2019.
But the COVID-19 pandemic caused life expectancy to drop by a full year, to just 77.8 years, in the first half of 2020, according to the tables accompanying a life expectancy estimates report.
The average life expectancy for 65-year-olds fell to 19.1 years, from 19.5 years in 2018.
Players in the insurance and pension communities are wondering whether the pandemic has mostly shortened the lives of frail people, leaving the country with a healthier population overall, or whether the pandemic has hurt the survivors’ health enough to shorten their life expectancy.
We’re interested in hearing from you. How do you see the U.S. life expectancy trends? Let me know what trends you’re spotting this week at abell@alm.com.
The buzz
Elizabeth Arias, Betzaida Tejada-Vera and Farida Ahmad, health scientists, National Center for Health Statistics
Provisional life expectancy at birth in the first half of 2020 was at its lowest level since 2006 for both the total population (77.8 years) and for males (75.1), and the lowest level since 2007 for females (80.5).
Life expectancy for the non-Hispanic Black population, 72.0, declined the most, and was the lowest estimate seen since 2001 (for the Black population regardless of Hispanic origin). The Hispanic population experienced the second-largest decline in life expectancy (79.9) reaching a level lower than what it was in 2006, the first year for which life expectancy estimates by Hispanic origin were produced (80.3). The levels observed for the non-Hispanic white population were last seen in 2005 for the white population (regardless of Hispanic origin).
Another consequence of the decreased life expectancy estimates observed in the first half of 2020 was a worsening of racial and ethnic mortality disparities. For example, the gap in life expectancy at birth between the non-Hispanic Black and white populations increased by 46% between 2019 and the first half of 2020 (from 4.1 to 6.0 years).
Regardless of Hispanic origin, life expectancy for the Black population has consistently been lower than that of the white population, but the gap between the two races had generally been narrowing since 1993, when it was 7.1. The gap of 6.0 observed in the first half of 2020 is the largest since 1998.
Conversely, the gap between the Hispanic and non-Hispanic white populations decreased by 37% between 2019 and the first half of 2020 (from 3.0 to 1.9 years). This indicates that the Hispanic population lost some of the mortality advantage it has evidenced since 2006 relative to the non-Hispanic white population, despite experiencing generally lower socioeconomic status,
The provisional life expectancy estimates presented in this report are subject to important limitations. First, they are based on deaths that occurred in the first six months of the year and do not reflect the entirety of the effects of the COVID-19 pandemic in 2020, or other changes in causes of death, such as the increases in provisional drug overdose deaths through early 2020.
There is seasonality in death patterns in any given year, with winter months typically seeing more deaths than summer months, and this is not accounted for in the data.
Second, the COVID-19 pandemic differentially affected certain geographic areas in the first half of 2020. The life table estimates may disproportionately represent mortality in those regions, which are more urban and have different demographic characteristics than areas affected by the pandemic in the latter part of the year. As a result, life expectancy at birth for the first half of 2020 may be underestimated since the populations more severely affected, Hispanic and non-Hispanic Black populations, are more likely to live in urban areas.
Sam Gutterman, consulting actuary
There are significant uncertainties concerning future U.S. longevity.
It is important to note that our population is quite heterogeneous, with a wide range of mortality risk characteristics and behaviors. It is the long-term aggregation of the effects of many of these factors that influence longevity. So, many observations regarding the “average” U.S. person fail to capture our true future longevity. For example, overall the last several decades those with “higher” socioeconomic characteristics have benefited from improved mortality, while those with “lower” socioeconomic characteristics have not.
Some people will continue to benefit from advances in medical diagnostics and treatment, while others will suffer from adverse changes in their behaviors and backgrounds. Longevity will continue to benefit from past reductions in smoking prevalence, while suffering from the seemingly never-ending trend toward increased obesity levels that has resulted from unhealthy nutrition and inadequate physical activity.
Unfortunately, the COVID-19 pandemic has attenuated alcohol consumption, overall body weight, and smoking, and slowed our success in addressing drug overdoses. I hope we can reverse some of these. And we don’t know what, if any, will be the health effects of the 10% of the American population who have survived COVID-19.
A key area of concern is the future of cardiovascular disease. Although over the last few decades we have greatly benefited from improvement in the identification and treatment of cardiovascular risk factors, such as blood pressure and cholesterol, there has been significantly less improvement in the 2010s.
It may be that we have seen the major benefits of our fight in this area and further advancement will be tougher to achieve. This may represent a major drag on future longevity improvement, as these diseases remain the most significant cause of death. Our success in these behavioral drivers of change in the cardiovascular area will determine whether our overall longevity continues to lengthen or we will remain stuck at our current level.
Yet we have plenty of potential to improve. We have wonderful health-related research capabilities, but we remain in the lower tier of longevity experience among developed countries. There is no reason why we can’t catch up. There are plenty of medical achievements ahead of us and room to improve our behaviors, but it will be a tough slog.
Nancy Bennett, senior life fellow, American Academy of Actuaries
Life insurance mortality is evaluated over a very long-term period. Products are priced using mortality and other assumptions for a 30-plus-year time horizon. Pricing assumptions include margins to cover varying mortality experience — i.e., products are priced assuming mortality will move up and down over the pricing period. The concern is when that fluctuation persists for an extended period and its impact can’t be absorbed in the margins.
Insurers’ mortality experience is different/better than general population mortality, as insurers use underwriting and risk selection criteria to create risk pools with better mortality experience than the general population.
Life insurance products include mortality guarantees where an insurer cannot charge the policy more than the mortality guarantee; actual charges are typically far less than the mortality guarantee.
If actual mortality experience turns out to be materially different than expected, some insurance products allow the insurer to make adjustments (e.g., dividends, cost of insurance charges). Note that raising prices comes with ramifications, so insurers are hesitant to do so unless necessary.
Mary Pat Campbell, life actuary and editor of Stump, a blog for actuaries
With respect to annuities in the United States, except for a very few companies, longevity doesn’t affect the financial performance of their annuity blocks of business — mainly because it’s mostly deferred annuities, and it’s the financial guarantees embedded in the products that drive those results and not much in the way of mortality.
But for life insurance, the mortality trends are concerning. The question is whether insurers are able to properly underwrite and price for the mortality risk, on both new policies as well as in-force coverage.
For products such as your standard annual renewable term group life insurance coverage through employment, that’s relatively simple. You can re-price for the next year’s business, so insurers should be able to be responsive and have only transitory issues with mortality trends.
For individual life insurance, even term life insurance, it’s more problematic, especially with the super-low interest rates. When valuation rates are so low, getting it wrong on estimating mortality levels is amplified, especially for longer-term term life and permanent products.
We already saw many companies having bad mortality experience (that is, worse mortality on their business than they had expected) on old blocks of universal life for many years before 2020.
Many companies increased cost of insurance (COI) charges on in-force business, which of course led to complaints (and lawsuits). I believe some of that was not only driven by population-level trends but also by the fact that lapse and surrender rates were very low for a very long time.
Low lapses are usually an indication that the policyholders can’t find better deals elsewhere, so that generally indicates the premiums are low relative to the benefits.
So there can be stress on these old blocks of business. While for some products insurers can increase charges to make up for deteriorating mortality experience, there are a lot of negative repercussions for that.
Other products, such as term life, have fewer levers to improve financial performance once they’ve been issued. Some insurers may seek to reduce their exposure by making deals to sell or reinsure in-force blocks of business.
READ MORE