Social Security has always mainly been a transfer system in which most or all workers’ payroll taxes immediately pay for benefits to retirees. The same is true for Medicare, though it also relies significantly on income taxes, which workers also largely pay. With the Social Security and Medicare trust funds now holding close to zero assets, it’s clear that any prior buildup of trust fund assets, such as during the peak working years of the huge baby boomer generation, was temporary and very small relative to the unfunded promises made.

Today, both systems face a significant shortfall in revenues needed to pay current beneficiaries at existing rates, not to mention future beneficiaries at higher rates. The shortfalls, which will increase over the coming decades, mainly stem from how Congress has scheduled these systems to take an ever-larger share of the nation’s income — almost indefinitely. Now, on the verge of a crisis, everyone agrees that reform is necessary. However, no reform proposal for Social Security (much less for Medicare) has addressed how large this transfer from younger to older generations should be.

Consider that, for the first time in the nation’s history and, with rare exception, in the history of civilization:

• People under age 65 are much more likely than older individuals to be in poverty or in lower-income groups.
• The government considers people “old” (eligible for what it originally called “Old Age Insurance” under Social Security) at age 62, when a man now has an average life expectancy of over 20 more years, a woman about a quarter century more, and the longer-living member of a couple about three decades. As healthcare continues improving, those numbers of years keep rising, and the Old Age Insurance program further evolves into a Middle and Old Age Insurance program.
• The total fertility rate for the country — the number of children expected to be born to the average woman — has long remained well below replacement level. And there is little evidence of when, or if, this decline might reverse or even stabilize.
• As a result of longer lifespans and fewer children, the Social Security worker-to-beneficiary ratio has fallen from about 4:1 in the early 1960s to 3:1 in 2009, then to 2.7:1 today, and is expected to decrease to around 2:1 (1.6:1 under pessimistic assumptions) within a half century.

Congress never designed Social Security to adjust to these changes fully. Without first examining how its fundamental structures and myths must adapt to this new reality, creating a rational and effective reform becomes nearly impossible. I keep getting approached by people about my assessment of different Social Security reform proposals. While I can comment on parts, I have no idea how much these proposals would shift resources from young to old, and frankly, I don’t think their advocates do either.

Why does every Social Security reform proposal I have seen avoid addressing this issue? As with many policy areas, we can easily fall into a time warp, thinking that a few tweaks will let us keep doing things mostly the same as before. We also assume we can always fix what is already a patchwork garment by clever sewing through the tears. 

When it comes to Social Security, reformers have long viewed it as an actuarial challenge: raising enough money to pay a set level of benefits for 75 years. Then, reformers add one tax fix here and one benefit fix there, mostly based on the current structure, to meet that actuarial objective. Naturally, Democrats focus on boosting benefit growth, while Republicans try to limit tax increases. But neither side clearly states the goals they want to achieve for both the elderly and the young.

Sometimes, moving to a new house is necessary when many features of the old one are no longer structurally sound. That’s not a bad thing. We can recreate the best parts of what we had, but now in a home that can better adapt to a changing environment.

To build the best possible future structure for Social Security, reformers must openly confront the fact that yesterday’s architects built today’s system on assumptions that are no longer valid — or never were. Among the flawed assumptions:

• That Congress must sustain a budget that consistently prioritizes older generations with increasingly higher transfers, no matter what other societal needs emerge.
• That one can set an ideal “annual replacement rate” of income for retirees without considering either how total lifetime benefits increase as people live longer or how the number of taxpayers available to fund that benefit is declining.
• That spousal and survivor benefits can (or ever could) be built fairly and efficiently around the stereotype of a male breadwinner who needs to contribute nothing additional to garner additional benefits for spouses and survivors.
• That continually providing more years of retirement support disproportionately benefits the wealthier and those who live longer.

Beyond Social Security, Congress created Medicare with few built-in budget limits on its growth at a time when total national health care costs were about 6 percent of GDP, compared to 18 percent today. Along the way, Congress effectively transferred its appropriations power to you and me once we become eligible for services (and demand more services), and to providers who can often set prices for what the government must pay.

Hidebound to old economic, cultural and demographic assumptions, both Congress and presidents also continue to treat the jerry-rigged combination of Social Security, Medicare, Medicaid, long-term care, Supplemental Security Income for the elderly poor, and pension subsidies as separate issues, as if the effectiveness of one bears no connection to the others. Likewise, they often handle decisions within each program — such as retirement age and funding for long-term care — as independent, when, in fact, at any given level of cost and progressivity, increasing resources for one goal decreases what is available for others.

There’s a similar problem with taxes. Reformers often treat each decision about what taxpayers should pay to support each system as separate. Then, within each system, they further obscure true costs. For example, with Medicare, Congress hides the costs from the public and itself by relying on multiple revenue sources, including payroll taxes, two sets of higher premiums for higher-income beneficiaries, income taxes, taxation of Social Security benefits and borrowing.

All this confusion about how Congress distributes benefits and taxes has led me to conclude that only an elderly reform, not just a reform of Social Security, can address these issues efficiently and fairly.

That said, we still face the core question of what constitutes a reasonable amount of net transfers (transfers minus taxes) from the young to support the old. Only after addressing that can we properly analyze how to allocate those transfers across benefit and tax programs.

Although this might seem like an abstract exercise, one practical approach is to compare the relative well-being of each age group over time. After all, transfer programs largely attempt to meet the needs of different groups. I also suggest that Social Security reformers originally adopted this perspective, though quite imperfectly, when they built the current system around the idea of a replacement wage. In reality, it wasn’t a genuine replacement wage but rather a method of transferring to a typical retiree an amount equal to about 40 percent of a typical worker’s wages.

• • •

I will elaborate in an upcoming column on my substack, The Government We Deserve. For now, I ask you to agree that you can’t design elderly or Social Security reform effectively without a way to assess how much the young should transfer to the old. Otherwise, Congress won’t understand what it is creating when it finally decides to assemble dozens of tax and benefit changes to hit some actuarial target. It won’t realize the extent to which it is driving and pre-ordaining broad social and cultural changes for both young and old. Finally, recognize that a fair and efficient solution today, when the old are generally wealthier than the young, will almost certainly differ from a past when the young were much wealthier than the old.