Jane the Actuary Retirement

Forbes post, “More Than An Insolvency Date: What Else To Know About The Social Security And Medicare Trustees’ Reports”

Originally published at Forbes.com on September 1, 2021.

 

I suspect that after years of reciting the same headline numbers, we tune them out:

The Social Security Old Age and Survivor’s Insurance Trust Fund will not be able to pay full scheduled benefits after the year 2033, one year earlier than forecast in last year’s report; when the Trust Fund is exhausted, it will only be able to pay 76% of benefits.

The Medicare Part A Trust Fund, which pays inpatient hospital benefits, will be fully funded through 2026 — no change from last year — and will be able to pay 91% of benefits at that point.

Each year we hear hand-wringing; each year those dates get one year closer. And, to be fair, this new date for Social Security is not as bad as various worst-case predictions earlier in the pandemic. But that’s not even the whole story.

Social Security

In the 2020 report (released in April and reflecting no impacts of Covid), the actuaries forecast that Social Security (OASI)’s cost rate would increase from 12.05% of taxable payroll in 2020 to 15.03% in 2040, decline slightly to 14.81% in 2045, and increase again, peaking at 16.19% in 2080. Income rates would rise only much more slowly, producing deficits of .88% (2020), 3.54% (2040), and 4.59% (2080).

This year, Social Security’s deficit is unusually high due to lower revenues and higher benefits: 1.75%. In 2040, the deficit climbs to 3.70% rather than 3.54%. In 2080, the deficit stands at 4.87% rather than 4.59%.

Put another way, if there were no Trust Fund accounting mechanism now, the OASI program would have been able to pay 93% of benefits. This would drop to 76% in 2035 – 2040 – 2045, then drop further to being able to pay 70% of benefits.

What’s more, this year, the actuaries changed several assumptions. They assume that by the year 2036, fertility rates will increase to 2.00 children per woman, an increase from the 2020 report’s assumption of 1.95. They also assume a long-term unemployment rate of 4.5% rather than 5%. At the same time, they calculate alternate projections with more pessimistic assumptions, including a continuingly low fertility rate (1.69), a higher rate of mortality improvement (that is, longer-lived recipients), a higher rate of unemployment (5.5%), and others. In these alternate calculations, the 2040 deficit becomes 6.47% rather than 3.7% (benefits 64% payable), and the 2080 deficit becomes 12.39% rather than 4.87% (benefits 50% payable).

Also consider that, at the moment, there are 2.7 workers for each Social Security recipient (2.8 in 2020). This is forecast to drop to 2.2 in 2040 and ultimately down to 2.1. But if the population trends are those of the pessimistic scenario, then that 2.1 would drop to 1.5 by the year 2080.

And, yes, once again, I continue to question the reasonability of the actuaries’ assumption with respect to fertility rates. In fact, in 2020, the actuaries had begun to reflect the ever-declining rate, which stood at 1.68 in 2019, even prior to the pandemic, by dropping the assumption from 2.0 to 1.95. This year, not only do they assume that every woman who deferred childbearing during the pandemic, will make up for it by having those “missing babies” in coming years, but they boost the fertility rate up from the 2020 reduction, back to 2.0, with no explanation offered!

Medicare

One would anticipate that Medicare’s finances would have been worsened considerably by expenses for covid patients in 2020, but, surprisingly, decreases in costs for non-covid patient care were greater than the increases in costs for covid patients, especially with respect to elective surgeries. However, the report itself acknowledges that the degree to which those expenditures will increase in the future as patients seek care that was foregone in the past, is highly uncertain.

In any case, projections in the future must estimate not only the same demographic and economic trends as for Social Security, but also changes in the cost of healthcare.

Taking into account only the Part A (HI) program, the only one with a “true” trust fund, the deficits are not particularly different in 2021 vs. 2020: a maximum deficit of 1.06% of payroll (remember there is no cap for Medicare) in 2045 vs. 1.08 in 2045 as of 2020, then declining to roughly half that. This works out to enough funds to pay 80% of scheduled benefit in 2045 and 91% at the end of the projection. But, again, in the high-cost alternate set of assumptions, Medicare would be able to pay only about 40% of benefits — and recall that isn’t anything that can be fixed with drug-cost negotiation or any similar promises, because these are hospital charges, the prices of which are already fixed at low levels by the government.

How urgently are fixes needed?

With respect to Medicare, the answer is, cynically, there’s not really much of a hurry. As I wrote back when the Biden administration introduced its 2022 budget, the administration’s willingness to fund an expansion of Medicare to younger ages simply through the use of general tax revenues rather than any dedicated payroll tax source, suggests that there is no fundamental reason why any part of Medicare at all needs this connection to the “Part A payroll tax.” Indeed, even with respect to Part A itself, there have already been transfers into the system with the CARES Act and similar legislation. There’s also no meaningful degree to which early action now will help us “save up” for expenses later — while we certainly do need a better way to run the system, one that improves health outcomes rather than paying blindly, but one that does not involve the degree of cost-shifting that occurs with Medicare’s reimbursement rates now, this has nothing to do with the trust fund itself.

With respect to Social Security, one aspect of the situation demands some kind of action: there is not even a legal mechanism for the Social Security Administration to respond to the depletion of the Trust Fund by deciding who does and doesn’t get benefits, or whether benefits would be reduced across the board or only for higher-income recipients. However, in principle, a “Social Security fix” could legislate some alternate funding source at any time.

At the same time, because government deficits are forecast to rise, year after year, and the demographic bulge of peak-earning Baby Boomers is long gone, there is no meaningful benefit to “saving up” for future benefits by trying to “rebuild” the Trust Fund.

Further, the “Social Security Reform” proposals of some individuals simply wish for the federal government to expand the benefits provided. That’s Biden’s proposal (which, incidentally, doesn’t even fully fund the system, suggesting a relative indifference to this question), among others.

But here’s where it does matter: it is not only my proposal, but, in various iterations, that of others, to re-invent Social Security to focus on providing a basic level of benefits while other legislation provides a framework for enhanced retirement savings by individuals, through retirement accounts or some sort of pooled system. This sort of new system would require a substantial phase-in period to enable workers to build up their balances. And this means, the sooner a reform happens, the better.

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Jane the Actuary Retirement

Forbes post, “Social Security Ending In 2023? No. But What Really Happens When The Trust Fund Is Emptied?”

Originally published at Forbes.com on August 27, 2020.

 

Earlier this week, the Social Security Administration’s chief actuary reported, in response to a question by Senate Democrats Chuck Schumer, Bernie Sanders, and Ron Wyden, that, if the dedicated payroll taxes to fund Social Security were removed with no replacement income source, the Disability Insurance Trust Fund would be depleted in the middle of 2021 and the Retirement/Survivors’ Trust Fund, in mid-2023.

This is no surprise — though the dates are sooner than my prior back-of-the-envelope calculation in a prior article. It’s fairly obvious that the key is the provision in the hypothetical, as specified by the Democrats, that there would be “no alternative source of revenue to replace the elimination of payroll taxes.” I have said regularly (and again in that that article) that I believe there would be a “Social Security fix” — legislation to simply direct general revenues to fill in the cap, whether it’s the 20% gap that has long been expected, or the full obligation in the event that the payroll tax were ended. I’ve further said that I could well see Congress legislating this as a “temporary fix,” over and over again, until the political stars align for a permanent change.

In other words, a headline such as NBC’s “Terminating payroll tax could end Social Security benefits in 2023, chief actuary warns” is highly misleading because it lacks that caveat: if there’s no other revenue source.

The notion that Trump would be able to corral enough votes to remove the payroll tax, but would not simultaneously ensure provision for alternative funding is nonsensical. In that respect, the Democrats’ request is simply an attempt to generate more headlines in their favor.

But that nonetheless suggests that there is a real risk that partisan rancor could cause the Trust Fund depletion to lead to a game of chicken, rather than to bipartisan, constructive solutions. So it’s time to address the question: what would actually happen?

The short answer is simple: when the Social Security Trust Fund is depleted, there will remain enough money to pay 80% of promised benefits.

This statistic is cited repeatedly, with the intent, by some, to generate urgency in solving the problem, while others use it to reassure their audience: “don’t worry, even if nothing’s done, the cut won’t really be that bad.”

But how exactly would benefits be cut?

There are, after all, a variety of options one might imagine.

Benefits could be cut in an across-the-board way: cut everything by 20%. Or, to be fairer, benefits could be given a haircut by resetting the maximum benefit at some lower amount. Benefits could be means-tested so that those with the most retirement income from other sources take the hardest hit.

But what does the law actually say? Turns out, the answer is “nothing.”

Literally.

Here’s the Congressional Research Service’s report, updated just this past July, “Social Security: What Would Happen If the Trust Funds Ran Out?”

“The Social Security Act specifies that benefit payments shall be made only from the trust funds (i.e., only from their accumulated bond holdings). Another law, the Antideficiency Act, prohibits government spending in excess of available funds. Consequently, if the Social Security trust funds become insolvent—that is, if current tax receipts and accumulated assets are not sufficient to pay the benefits to which people are entitled—the law effectively prohibits full Social Security benefits from being paid on time.

“The Social Security Act states that every individual who meets program eligibility requirements is entitled to benefits. Social Security is an entitlement program, which means that the federal government is legally obligated to pay Social Security benefits to all those who are eligible for them as set forth in the statute. If the federal government fails to pay the benefits stipulated by law, beneficiaries could take legal action. Insolvency would not relieve the government of its obligation to provide benefits.

“The Antideficiency Act prohibits government agencies from paying for benefits, goods, or services beyond the limit authorized in law for such payments. The authorized limit in law for Social Security benefits is the balance of the trust fund. The Social Security Act does not stipulate what would happen to benefit payments if the trust funds ran out. As a result, either full benefit checks may be paid on a delayed schedule or reduced benefits would be paid on time.”

But does the Social Security Administration have the authority to made decisions about how to pay out benefits?

To operate within the strict parameters of the law, the administration would be obliged to simply delay benefit payments, creating a backlog that grows . . . and grows . . . and grows — because, however much the original narrative may have been that Social Security deficits are temporary, due to Baby Boomer retirements, that’s not at all the case; we will never again be able to cover Social Security costs through FICA taxes, and the deficit will only get worse over time. This means that retirees would get “full” checks but only 9 or 10 in a year.

Could the Social Security Administration cut benefits, to manage this backlog, so that everyone gets 80% checks paid on time? For further clarity, I checked in with Marc Goldwein, head of policy at the non-partisan Committee for a Responsible Federal Budget, who confirmed that there is no consensus on this point; some experts say “yes,” some say “no.”

What it comes down to then, is this: it may indeed be the case that, come 2035, or 2031 or 2027, Republicans and Democrats will both recognize that they have to sort out their differences so as to not leave Social Security in the lurch. But if they insist on “winning” rather than compromising, the outcome could be quite chaotic indeed.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Jane the Actuary Retirement

Forbes post, “What If We Stopped Worrying About The Social Security Trust Fund?”

Originally published at Forbes.com on June 11, 2019.

 

Readers, I’ll start with a reminder:

I believe that Social Security doesn’t need just a little bit of renovation, taking out the old wallpaper and replacing the carpet with hardwood, but needs a whole-house gutting, in which we swap out, for future accruals, the current clunky bendpoint-based formula for a flat anti-poverty benefit paired with a mandatory account-based funded system on second-tranche income.  (Democrats are already proposing, via the “Social Security 2100 Act,” a much-increased minimum benefit, and various states are implementing state-sponsored mandatory auto-enrollment IRA programs, but I am convinced that the only way to implement a reform that will have widespread bipartisan support is by incorporating a funded system, with smoothing and pooling, into the Social Security system in place of the current structure, for middle-income pay-replacement.)

I acknowledge, however, that, to date, I have succeeded in persuading, near as I can tell, 0% of Congress.

That being said, so long as our imagination is limited to new coats of paint, I’m increasingly inclined to believe that we should simply discard the idea of caring about the end point of the existing Trust Fund.

If, on the one hand, life and health expectancy has increased such that it’s reasonable and appropriate to increase the retirement age, then we should do so regardless of what other decisions are made.  (My personal preference would be to legislate a method for adjusting the retirement age on a regular basis at that age at which a fixed percentage of workers, say, 1/4 or 1/3 of the group, would otherwise need to retire for disability/health reasons.)

If, on the other hand, the consensus is that we need some combination of tax hikes and earnings-cap adjustments, and if at the same time, there’s a majority belief that tax revenue needs to be boosted in any case, why do we need to play games with a “Trust Fund”?  Why not simply adjust taxes to the level needed to remedy deficits and fund whatever other spending the majority wants, in ways that are sustainable in the long term but recognize that demographic challenges will grow over time?  For that matter, why, if we want a benefit formula in which the wealthy subsidize the poor, rather than a “pay your own way” contribution, why should we limit taxes to wage income?

And, again, recall that of far greater significance than the Trust Fund is the old-age dependency ratio; back a year ago I referred readers to a Brookings statistics that total federal spending on the elderly is projected to increase from today’s 20.5% of GDP to 29.4% of GDP in 2046 — not 29.4% of government spending but of the total economic output.   How we cope with this aging future matters more than an arbitrary Trust Fund Depletion Date.

Consider the Disability Trust Fund.  In the 2016 Trustee’s report, this fund was projected to be depleted as soon as 2023.  In 2017, that number moved to 2028, in 2018, to 2032, and in the most current report to 2052.  Is Social Security Disability Insurance “fixed,” then?  No, of course not.  The program has widely-acknowledged problems, as many genuinely disabled Americans find themselves obliged to hire lawyers (and pay their contingency fees) in order to navigate the system, and others succeed in collecting benefits despite a genuine ability to work; hence, this stretching out of the fund depletion date is the result not of American’s improved health so much as the improved economy keeping more people in the workforce who would otherwise be deemed unable to work.  Other acknowledged shortcomings of the system include the lack of partial disablement structure to keep Americans in the workforce if they are able to work on a partial basis, and an insufficient return-to-work program.  (See, for instance, “16 Reforms to Improve the Solvency and Integrity of Social Security Disability Insurance,” a report at the Heritage Foundation or “Disability insurance: A crisis ends, but problems persist,” at the Brookings Institute, for recent commentary.)  For that matter, a report in yesterday’s Washington Post about a proposed disability change, in which blue-collar workers who are not proficient in English would no longer qualify for disability on the principle that they can’t transition to a desk job, signals a fundamental flaw in a system of all-or-nothing, which leaves no room for temporary benefits while an individual leans a new occupation, be that skills for a desk job, or English language instruction.

And none of these issues have anything to do with the depletion of the Trust Fund — but have we so conditioned ourselves to think of the depletion date as the target that we can’t think sensibly about more fundamental reform?  Certainly, the drumbeat of “we have to reform Social Security Disability” has diminished quite considerably as that date was moved further and further into the future.

And the same is true of “regular” Social Security, that is, the Old Age and Survivors’ program.  Yes, we must craft a benefit design and funding structure that is sustainable in the long term, and which takes into account anticipated life and health-expectancy improvements and expected changes in the long-term in fertility rates, economic development, and the like.  But to take as our marker the Trust Fund depletion date, or set as our objective replenishing the Fund?  Let’s not.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Jane the Actuary Retirement

Forbes Post, “The Social Security Trust Fund Is Real – But So What?”

Originally published at Forbes.com on May 5, 2018.

 

Stop me if you’ve heard this one before:  “Social Security would be doing just fine if the thieving politicians hadn’t stolen all the money from the Trust Fund that we paid in to earn our benefit checks.”

Or maybe this:  “The Trust Fund is just an accounting gimmick, nothing more, and Social Security is broke.”

Who’s right?

The Trust Fund is real.

Well, sort of.

The Social Security Administration does indeed invest its surpluses, that is, the revenues from FICA taxes, taxes on Social Security benefits, and interest credited to the fund, less benefits paid out, into government bonds.  And if you or I, or, say, a pension fund, happened to exist in government bonds, we wouldn’t consider that investment to be “fake,” or the money to have been “stolen” from us.  It’s real money, and we’d have a real right to that money.  In the same fashion, the Trust Fund will redeem its bonds when it begins to run deficits.

But that’s not the end of the story.

Consider that the trust fund is not a matter of “us” saving for “our retirement.”  There was, to be sure, a real element of building up surpluses during the early years of the program, though not to the degree that would truly make it an advance-funded program, rather than only partially-so.  In any event, the Trust Fund was virtually depleted in 1983, and the system would have been unable to pay full benefits but for FICA contribution increases beginning in 1977 and accelerated in a 1983 bipartisan reform bill, which also raised the retirement age to 67 and otherwise stabilized the system’s finances.   The funds which have built up since then are the moderate excess of revenues over payouts following the tax hike, not a true pre-funding as you’d see, for example, in a private-sector pension plan.  Its function is really just, in principle, to smooth out spending over time.

The trouble is, though, that one can outline what the Trust Fund “is” in terms of accounting and financing, but people tend to look at this in a moral sense.  The Trust Fund is the embodiment of American workers’ conviction that, having paid taxes during their working lifetime, they have a moral right to their Social Security benefits, or, more generally, to a retirement free of financial worry.  And this is not the case.

Consider this alternative:  what if, in the 1983 Social Security reform, rather than building up surpluses, Congress had decided that any surpluses would be become general tax revenues and any deficits would be paid directly through tax revenue?  Functionally, there would have been little difference, other than bookkeeping, between building up a fund, nominally, that is immediately funneled into government spending, and doing so directly, and between bonds being redeemed, in the future, requiring new borrowing to fund the redemptions (or running a budgetary surplus — in some alternate universe, anyway), or just borrowing directly.

Alternately, what if Congress had simply decided that FICA taxes would vary each year, determined by the projected Social Security payouts each year?  We would not be discussing the depletion of a fund, but instead, perhaps, would be complaining at the prospect of our FICA taxes growing ever higher.

What would the economy of the United States have looked like in the past several decades had there not been FICA surpluses used to buy (or “buy” with scare-quotes if you like) government bonds?  Consider that the Social Security Trust Fund (in combination with the Disability Trust Fund), at the end of 2016, “owned” 13% of the total National Debt (the link includes a detailed breakdown of what entities own what portion of U.S. debt).  Did the availability of the Trust Fund as a debt-purchaser, help hold down interest rates, keep government borrowing affordable, and keep the deficit lower than it otherwise would have been?  Or did this simply enable Congress to defer dealing with deficits when they might otherwise have been motivated to make hard decisions?

Or consider our Neighbors to the North.  Canada, after all, has a Trust Fund, but the nature of the fund is radically different:  it is a real investment fund, holding a wide variety of assets,  including private equity and real estate holdings (they fully or partially own Petco, Univision, and Neiman Marcus, for instance).  Their long-term planned asset mix is 55% equities, 20% fixed income securities (largely government bonds) and 25% real estate.  (Readers can learn more at the Canada Pension Plan Investment Board website.)  In addition to providing a higher rate of return over time than the interest credits of the U.S. Social Security Trust Fund, the very nature of the Canadian fund is wholly independent of the government.  What’s more, although the Canada Pension Plan has historically been more-or-less pay-as-you-go just as in the U.S., they have actually just recently implemented a benefit increase, which is being phased in slowly enough to be wholly pre-funded by a payroll tax increase.

The surplus that generated the Trust Fund were a missed opportunity.

To be fair, there have been worries about the prospect of the government of the United States managing a sovereign wealth-type fund of such a massive size.  Could an investment board truly make decisions impartially?  Would the government be too heavy-handed, attempt to micromanage the companies in which it invested — for example, by monitoring executives’ salaries for “fairness” or requiring a sufficient number of female or ethnic-minority board members?  Maybe.  But there would have been an alternative — the time would have definitely been ripe for an alternate Social Security system, in which the pre-funded component from those surpluses could have been in the form of individual accounts or pooled but nongovernmental funds (hmmm . . . where have I heard that before?).  Such a system would have allowed for the buildup of real advance funding for retirement, rather than leaving us worried about the future.

But the “Real-ness” of Trust Fund is a bit academic.

The bottom line is that whether the Trust Fund is “real” or just a fiction on paper, in the end, doesn’t matter.  Whether the Trust Fund uses its assets to pay retirees, and the federal government has to borrow, to pay back that debt, or whether the government has to pay those benefits directly, it’s still the case that money has to be found — and the amount of money which will have to be found, for retirement benefits, Medicare, and other expenses, is forecast to grow dramatically.  A January paper from the Brookings Institute provides some very sobering numbers:  due to the aging of the American population, federal spending on the elderly is forecast to grow from the current (2017) level of 20.5% of GDP, up to 29.4% in 2046 — and that’s not 29.4% of government spending, but 29.4% of our total economic output.  And this isn’t just a temporary “hump” due to the Baby Boom.  The author states:

Although we often talk about aging as arising from the retirement of the baby boomers, that is somewhat misleading. The retirement of the baby boomers represents the beginning of a permanent transition to an older population, reflecting the fall in the fertility rate that occurred after the baby boom and continued increases in life expectancy. Because aging is not a temporary phenomenon, we can’t simply smooth through it by borrowing. Instead, it is clear that population aging will eventually require significant adjustments in fiscal policy—either cuts in spending, increases in taxes, or, most likely, some combination of the two.

What should the policy response be to future impact of an aging population?  The paper acknowledges that such forecasting is uncertain, and offers various options but does not promise any easy solution — because there is no easy solution on offer.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.