The editors of Bloomberg recently published a series of editorials on America’s impending “retirement crisis,” which they assert could “under reasonable assumptions … overwhelm state budgets across the country” as impoverished seniors increasingly rely on government handouts to survive. Scary stuff, if true. In fact, the assumptions driving these frigthening claims are anything but reasonable. And projections from nonpartisan government agencies tell almost the opposite story: poverty in old age is falling, not rising, as America’s seniors grow ever-richer.
Seemingly everyone believes that Americans are dramatically undersaving for retirement. Too few employees are participating in 401(k)s; if they do participate, they don’t contribute enough; and once they retire, Americans lack the financial acumen to manage their savings over a retirement that could last for decades.
But assuming the critics are right, what happens then? The answer, according to Bloomberg, is that government welfare programs will bear the burden. Supplemental Security Income (SSI) provides cash payments to the poor elderly; Medicaid provides health coverage; the Supplemental Nutrition Assistance Plan (SNAP) provides “food stamps”; home energy subsidies help with heat and power; and other programs vary by state or local government discretion. If the retirement crisis comes, poor retirees could cost the government a great deal of money.
And that is what Bloomberg claims will happen, based upon an AAR
But, as with many, many media stories detailing our impending “retirement crisis,” these conclusions fall apart once we look at the details. The New Jersey study starts with data on household income and assets for the year 2008, drawn from the Survey of Income and Program Participation (SIPP). From household earnings the study estimates future Social Security benefits and from household savings it estimates the assets retirees can draw down in old age. Together, Social Security and asset income are assumed to make up total income in retirement.
All that sounds reasonable, but a close watcher can tell there is a problem from the very outset: welfare costs for New Jersey seniors don’t gradually rise from the actual value of $404 million in 2016 to $7 billion or more in 2030. Rather, from the actual 2016 values to the study’s first projected values for 2020, welfare costs rise from $404 million to $2.2 billion, a more than five-fold increase in just four years. The obvious explanation for that increase is that the study’s model underestimates retirees’ incomes and therefore overestimates costs for means-tested welfare programs.
I can suggest where those problems might arise. First, the AARP study estimates Social Security benefits based upon the survey respondents’ current earnings. If their earnings continue to rise, which is likely, their Social Security benefit also would be higher.
Second, the study ignores income drawn from traditional “defined benefit” pensions. Not many private sector workers are accruing new benefits under traditional pensions, but a larger number did so in the past and can expect some benefit in retirement. Moreover, nearly all government employees have traditional pensions. According to the Public Plans Database, over 339,000 New Jersey residents received public sector pensions in 2020, with an average annual benefit topping $33,000. That’s a lot of money to leave out of the study’s estimates.
And third, while the study assumes that the household assets measured in 2008 continue to earn 4 percent annual interest, the study assume that no additional retirement savings taking place after 2008. An individual turning 65 in 2035 would have been only age 38 in 2008, with the bulk of their saving years still to occur.
In sum, Bloomberg bases its conclusion that “America’s Retirement Crisis Is a Financial Crisis Too” based on a study that underestimates Social Security benefits, omits generous public employee pensions, and assumes that households save nothing for retirement after the year 2008. In my opinion, these limitations render the study’s conclusions essentially meaningless.
However, two nonpartisan government agencies have made estimates that can help. The Social Security Administration’s Office of Policy uses a sophisticated microsimulation model to project the incomes of future retirees in great detail. SSA estimates that in 2022 about five percent of Americans 65 and over had incomes below the poverty threshold. But by 2050, SSA projects, only four percent of retirees will be below the poverty line. Similarly, SSA records that in 2022, 3.9 percent of seniors were eligible for Supplemental Security Income (SSI) benefits, a means-tested program for the poor. SSA projects, however, that by 2045 only 3.3 percent of seniors will have incomes and assets low enough to qualify for SSI. Likewise, the Congressional Budget Office projects eligibility for the Medicaid aged program, which provides health coverage to low-income seniors. From 2020 to 2031 the CBO projects that the share of the over-age 65 population covered by Medicare will decline from 12.8 percent to only 9.7 percent.
All of this is part of a longer-term trend of rising retiree incomes and declining poverty in old age. It is regrettable that a much relied-upon publication like Bloomberg would treat rising old-age poverty as a fact when the best nonpartisan estimates tell us just the opposite. Americans’ retirement security is far too important an issue for the facts to be treated so casually.