Public Employees Are Living Longer Than Previously Assumed, New Report Finds

Many state and local governments continue to grapple with exorbitant unfunded pension liabilities, with underperforming assets and insufficient employer contributions being the major culprits. Another contributing factor that often gets overlooked, however, is the use of outdated mortality rate assumptions. A recent study by the Society of Actuaries (SOA)—which for the first time specifically focused on public pension plans—confirmed a long-standing belief that public workers in certain professions do live longer than the rest. This could potentially portend higher costs of pre-funding benefits for many small to mid-sized public pension plans that rely on private-sector mortality tables for their liability projections.

“From that [RP-2014 Mortality Tables] study we noted differences in public plan mortality. That led to a request for data from public plans,” said Dale Hall, managing director of research at the Society of Actuaries.

Five years in the making, the mortality and longevity experience study conducted by SOA—which collected data from 78 public pension plans—sheds light on a debate over the life expectancy of public vs. private employees. The results suggest that male and female teachers, as well as female general government employees who reach age 65, are expected, on average, to live longer than previously assumed.

To be fair, most people, in general, are living longer. For example, in Japan, the number of centenarians is expected to skyrocket from 62,000 in 2015 to 532,000 by 2050. Japan finances its public pension almost exclusively on a pay-as-you-go basis, which means current workers need to subsidize an increasing number of retired members.

This is in line with other troubling trends. Per Census Bureau projections, by 2030 more than 20 percent of U.S. population will be 65-years-of-age or older. And as a result of low fertility rates and the expanding elderly population, the financial sustainability of large social safety nets—like a pay-as-you-go Social Security system in the U.S.—is steadily deteriorating.

It is true, however, that an aging population per se might not directly impact state and local pensions in the U.S., an overwhelming majority of which are pre-funded defined benefit (DB) pension plans. That is because every current public employee participating in a DB plan is not supposed to subsidize the increasing pool of current retirees, but rather only pays toward his or her own retirement benefits over time.

However, longer-than-projected lifespans do raise the cost of pre-funding such lifelong pension benefits for both government employees and government sponsors, which typically use private-sector mortality tables. And costs have nowhere to go but up.

A recent Society of Actuaries study of private-sector pension plans shows that members live longer than the general public but shorter than many government workers: men in these systems live on average to about 85.6 years old, and women to slightly more than 87.6 years old.

All the while, according to the SOA public pension longevity study, female government workers in the general public-pension plans lived to 88.5, while male government workers lived up to 85.5. Educators came out on top. Female and male teachers who reach age 65 are expected, on average, to live to 90 years and 87.7 years, respectively. Even longevity expectations for public safety officers were slightly improved, with males and females expected to live up to 85.3 and 87.7 years, respectively (see Table 1 below).

Table 1. Life Expectancy of Public Employees Who Reach Age 65

* All base mortality rates are projected with MP-2017 projection scales.

Source: Society of Actuaries, Pub-2010 Public Retirement Plans Mortality Tables

The percentage change indicates by how much, if revised, mortality rates of public teachers, public safety officers, and other public employees would differ from the private-sector 2014-based mortality rates (so-called RPH-2006 Mortality Table) used by many public pension plans, such as Arkansas TRS.*

That is not to say that these revisions are relevant to all publicly funded retirement systems. For example, some large public pension plans, such as the California Public Employees’ Retirement System, CalPERS, already rely on internal analysis of members/retirees for mortality and longevity estimates. However, many other small to mid-size pension plans use private-sector mortality tables to estimate actuarial liabilities and contributions necessary to fully fund promised benefits.

As an example, the Teachers Retirement System of Georgia (TRS) —which until 2010 was using 1994-based mortality assumptions—accrued over $390 million in unfunded liabilities from 1997 to 2017 solely because retired teachers lived longer, thereby receiving more pension checks than the plan originally anticipated. This is an integral part of the plan’s overall demographic experience, something highlighted by colleague Jen Sidorova in a recent report by the Georgia Public Policy Foundation and the Pension Integrity Project at Reason Foundation.

Other examples of how relevant these assumptions might be for pension costs include:

• New York, where—after the state’s retirement system updated its mortality tables in 2015—localities found themselves in need of increasing their pension contributions from 14.2 percent of salaries to 18.2 percent, according to S&P.
• Several communities in Illinois, which in 2014 discovered that they had been using mortality tables from as far back as 1971, grossly underestimating pension costs, which resulted in about a 20 percent average increases in costs.
• Managers of Milwaukee County’s Employees’ Retirement System (MCERS), which in 2013 discovered that as many as 850 county retirees had been either underpaid or overpaid by considerable amounts because of a misapplied mortality table.

With longer life expectancies, the case to further de-risk pension asset allocations and improve liquidity becomes even more pronounced. That’s because, as the average age among both active members and retirees increases, so too must the investment risk tolerance to secure soon-to-be-made benefit payouts.

As the SOA study shows, public employees, on average, have lower mortality rates and longer life expectancies than private-sector pension plan members, which should put some government sponsors on notice for higher anticipated costs. And periodically fine-tuning inaccurate actuarial assumptions, including mortality rates, is a prudent approach that would help public pension systems stay on course for the full funding of their promised benefits.

[*Note: If, however, the revised mortality rates for public employees in the SOA study were to be compared not to the 2014-based, but other commonly used 2000-based mortality rates (so-called RP-2000 Mortality Table)—Arkansas PERS pension plan—instead, then life expectancies of both male and female public safety employees, as well as male general government employees aged 65 should be revised downwards, not upwards.]