June 25, 2018

Consequences of Financing Paid Parental Leave Using Social Security

Social Security is the largest government program. In FY 2017 it paid over $939 billion—4.9 percent of GDP—in benefits to 62 million Americans (45 million of whom are retired). Unfortunately, the program is not on sound financial footing, as 17 percent of currently scheduled benefits lack funding, and the two Social Security trust funds combined face depletion in 2034. Upon trust fund depletion, program tax revenue will only be able to pay 77 percent of benefits. The financial troubles facing the Social Security program are among the many reasons why a new proposal to use Social Security as a way to extend paid family leave to new parents is a bad idea.

The plan proposed by Kristin Shapiro of the Independent Women’s Forum in her recent paper “A Budget-Neutral Approach to Parental Leave” appears at first to be simple and elegant.

Consider a 26-year-old new mother with five years of work experience earning $31,100 per year. Under this plan, she could receive 12 weeks of paid leave at a rate of close to $300 per week, or approximately $3,600 over the 12-week period. In exchange, in the future, her eligibility to claim Social Security retirement benefits would be delayed by about six weeks.

This is a simple, stylized example. The plan lacks necessary details for a more complete analysis. In a recent piece in the Wall Street Journal, Andrew Biggs of the American Enterprise Institute and Kristin Shapiro provide additional numbers on the potential cost of the proposed paid parental leave policy:

The cost is low because parental-leave benefits claimed early in life would be low relative to retirement benefits claimed later, as earnings typically rise considerably from one’s 20s to one’s early 60s. . . . Consider an average woman, who enters the workforce at 21 and has her first child at 26. At 25 she would have a salary of about $31,100, according to Social Security Administration data. Using the same formula used to calculate Social Security disability benefits, she would be eligible for a Social Security parental-leave benefit of $1,175 a month, equal to 45% of her earnings at 25. . . . Because of Social Security’s progressive benefit formula, lower-income workers would receive a higher benefit relative to their earnings.

To the designer’s credit, this plan doesn’t involve mandating employers to provide paid leave to their employees, and it piggybacks on an existing entitlement program rather than starting a new one from scratch. Its apparent simplicity and purpose, topped with the claim that it wouldn’t add to the deficit, has already attracted support. It was taken up by both conservative and moderate members of the Republican congressional conferences, and it is gaining some traction among conservative groups.

Unfortunately, while the plan lacks many details, there are already considerable reasons to oppose it. The first question Americans should ask is whether paid family leave is best provided through the government and whether the private sector is providing, or could provide, such a benefit. Data from the Census Bureau’s Survey of Income and Program Participation (SIPP), the US Department of Labor’s FMLA Worksite and Employee Surveys, the Census Bureau’s Current Population Survey (CPS), and the National Survey of Working Mothers show that even without the government mandating or paying for a parental leave benefit, between 45 percent and 63 percent of women report already having access to paid leave. This should come as no surprise, since academic studies show that companies benefit from providing this type of benefit to their workers—adding satisfaction, as well as lowering turnover rates and raising productivity. When businesses can afford to offer paid family leave, they increasingly do. It is unclear that there would be net societal benefit in shifting the costs of paid parental leave from the workplaces that benefit from it to federal taxpayers.

Setting these general concerns aside, this paper will focus on a more specific problem with the plan: the use of the Social Security trust funds to finance paid parental leave. We will discuss how this proposal is an inappropriate expansion of the mission of the program, how it disrupts the link between work and benefits, and how it will almost certainly not be budget neutral, among other issues. Rarely do the finances of such programs conform to their designers’ original intentions. If a paid leave program were to be added to Social Security, over time we could expect that its eligibility criteria would be expanded (as they have been with other entitlement programs) to include other paid leave options besides caring for a newborn, such as caring for aging parents or a spouse. The duration of the benefit would likely be extended over time as well, as paid leave benefits in other countries have been. The final transformation of the new benefit may happen when advocates for beneficiaries demand that it be paid for with general government revenue rather than delayed retirement. In sum, it is reasonable to expect that this new entitlement would not long remain a budget-neutral modification of existing mandatory spending obligations, but would eventually become an entirely new spending obligation, with all of the costs and market distortions that such entitlements typically create.

Inappropriate Expansion of the Role of Social Security

Using Social Security as a way to provide paid leave would be a significant expansion in the types of benefits that Social Security pays for, and one that departs significantly from the program’s historical mission.

The historical mission of Social Security has been to provide “a comprehensive package of protection” against income loss as the result of a permanent departure from the workforce, consequent to a primary household earner’s retirement or death. This system was originally designed as a social insurance program to cover low-probability and high-cost events (such as living well beyond the average life expectancy or losing income if a working spouse dies). Later the program was expanded to include disability benefits. That expansion took place only after a very protracted debate during the 1950s and after proponents of the expansion gave multiple assurances and procedural guarantees that the establishment of disability insurance benefits would not undermine the funding of the retirement program. As part of this process, and to fulfill this requirement, the Disability Insurance system was established with a separate trust fund and a separate payroll tax. As Senator Walter George (D-GA) stated during floor debate,

The moneys for disabled persons will not be commingled in any way with the funds for old-age insurance or for widows and spouses. The contribution income and the disbursements for disability payments will be kept completely distinct and separate. In this way the cost of disability benefits always will be definitely known and the costs always will be shown separately . . . a separate tax is to be levied to build up a fund which can be easily policed, which can never encroach upon the fund for widows, and for those who reach age 65, and for children and other beneficiaries.

Even still, disability benefits were only provided for those facing long-term income loss owing to a disabling condition. None of those safeguards ground the current paid leave proposal; instead, it seems the Old-Age and Survivors Insurance (OASI) Trust Fund would be required to finance a new kind of benefit entirely—one that does not align with the original intent of the program. Given that the OASI trust fund is already significantly underfunded, it seems imprudent to start tapping those inadequate funds to pay for a new category of benefits.

To elaborate on the previous point, paid leave is a fundamentally different kind of benefit than the other benefits Social Security pays for. Although people do risk income interruptions when they leave the workforce to care for children, having children doesn’t permanently destroy one’s ability to earn income, now or in the future. Social Security is designed instead to protect against lasting income loss—because someone either died, retired, or became disabled. For better or worse, in order to receive disability benefits, applicants must show that their disabilities are an ongoing condition; Social Security explicitly says that “no benefits are payable for partial disability or for short-term disability.” In other words, Social Security has never been intended to provide temporary “tiding-over” benefits, unlike, for example, unemployment insurance.

Paid Leave Would End the Relationship between Work and Benefit

Social Security is a program with many problems, including its projected insolvency. However, one of the program’s relative virtues is that no workers get benefits that they haven’t at least in some way earned with their payroll tax contributions. On the retirement side of the program, one must work and contribute payroll taxes for at least 40 quarters (10 years) to be eligible for benefits.

Disability is more complex, and one could argue that its benefits aren’t adequately tied to contributions paid. However, its design is defensible from an insurance perspective since disability can strike at any time and cause a permanent loss of income. Yet even on Social Security’s disability side, one must work and contribute for a few years before qualifying (6 quarters of tax contributions are required to become eligible before the age of 24; otherwise, 20 quarters are required over a worker’s previous 10 years). And although some young workers receive disability benefits that their own contributions have not yet adequately funded, it is far more common for workers to receive disability benefits only much later in their careers, after several years of contributing payroll taxes to the system.

By contrast, paid parental leave is ill suited for this type of contributory funding system. Most people tend to have children toward the beginning of their careers rather than toward the end, and well before their contributions, even in the aggregate, are sufficient to fund benefits. Under a paid parental leave policy, many people will likely have children long before they have contributed payroll taxes for 20, let alone 40, quarters of paid employment.

To address this issue, the proponents of the paid leave system would require that parents, no matter their age, need to have worked at least four quarters (one year) in their lifetime, including in at least two of the last four quarters preceding the birth of their child, to become eligible for Social Security parental benefits.

This is a serious weakening of the work-benefit relationship that currently exists in Social Security. Moreover, it is easy to anticipate that even these minimal work requirements will eventually be relaxed or even eliminated to provide additional assistance to lower-income parents without established labor force attachment. Establishing a new benefit that is not premised primarily upon prior employment-based contributions will likely render it inevitable that the benefit is ultimately conceived and provided solely on the basis of need, rather than earned through work. While there is a place in government for purely need-based assistance to poor families, Social Security is not that place.

A Not-So-Budget-Neutral Proposal

Advocates for the paid leave proposal express an intention that this new benefit be budget neutral and not add to the net cost of Social Security. As Shapiro writes, “The proposed program would be structured to be self-funding. In return for receiving parental benefits, new parents would agree to defer their collection of Social Security benefits upon retirement for the period of time necessary to offset the cost of their parental benefits.”

For instance, if a 25-year-old had a baby and claimed paid leave benefits in 2018, she would effectively be using her 2060 old-age retirement benefits. That’s assuming she wants to retire at 67. She would make up for her paid leave benefits in 2060—over 40 years later—by delaying her retirement benefits by six weeks. For each additional child, she would forgo an additional six weeks of retirement benefits, and so on and so forth.

It seems simple enough. But while in theory this paid family leave is potentially budget neutral over the lifetime of the individual Social Security beneficiary, it would nevertheless add to Social Security’s solvency problems for several years, if not decades. Social Security has been running a cash flow deficit since 2010—one that is currently projected to be permanent. Starting from the implementation of any new paid parental leave program, additional borrowing will be required to pay for the paid leave benefits claimed by parents, on top of all Social Security benefits currently going to retirees. According to Shapiro, assuming a 25 percent take-up rate by eligible parents, that additional cost would add up to $7 billion per year.

These additional benefit outlays will accelerate the depletion of the trust funds, which the 2017 Social Security Trustees report already projects will happen in 2034, and at which time Social Security benefits would be reduced by roughly one-quarter under current law. By 2034, the federal budget deficit is estimated to be $2.4 trillion (6.5 percent of GDP), and the public debt to be $38.4 trillion (105 percent of GDP), to which the costs of this additional benefit would be added.

In other words, using Social Security to pay for family leave means adding fiscal pressure to an already unsteady system in a highly indebted budgetary environment. Lawmakers’ six options for dealing with the Social Security shortfall would all become more severe: sharp reductions of benefits, significant delays in the age of retirement eligibility, steep increases of the payroll tax rate above its current 12.4 percent level, lifting the $128,400 income cap subjected to the tax, bailing out the system with general revenues and eroding its status as an earned benefit, or a mix of the five.

Designed to be part of a multifaceted system for income replacement in retirement, Social Security benefits, along with employer pensions and individual savings, form the so-called “three-legged stool” of retirement. Social Security alone cannot provide adequate retirement income for all, and it currently lacks adequate funding for the benefits it promises.

It would be especially unwise to establish the principle that Social Security’s future retirement benefits can be borrowed against to finance the needs of the young in the present. In the first place, promises of Social Security benefits are funded only by future taxpayers; there is no storehouse of personal Social Security savings for young workers to draw from. And even if there were such a storehouse, there would be no logical basis for limiting access to it only for paid parental leave; the door would be open to unlimited alternative uses for such advance payments, from down payments on a home to paying off student loans. Indeed, some of these expansions are already being proposed.

Nor should Americans take it as a given that the proposal would actually be self-funding, even over the lifetime of a beneficiary. This would be a “benefits now, funding later” program. There is no fail-safe mechanism for ensuring that lawmakers follow through on the benefit offsets decades hence. In some cases, follow-through may even be impossible, such as for those who receive benefits up front but then remain out of the workforce as stay-at-home parents. According to the Bureau of Labor Statistics, about 35 percent of women with young children do not participate in the labor force. The program would also fail to be self-funding in instances of workers who later become disabled or drop out of the workforce and thus have no retirement benefits to delay.

It is worth adding that government paid leave programs around the world have a bad track record. As Vanessa Brown Calder reminds us,

In 1989 Larry Summers wrote about government-mandated paid leave, “There is no sense in which benefits become ‘free’ just because the government mandates employers offer them to workers.” And in 1994 Jonathan Gruber reported women’s wages were reduced to reflect the cost of benefit mandates. Gruber estimated that the shift in cost was around “the order of 100 percent.”

Under the current paid leave proposal, the government wouldn’t mandate or provide net funding for the benefit—at least at first. Yet it is naïve to think that no new costs will emerge as such a program inevitably evolves. Potential adverse consequences of these costs include unexpected burdens and disruptions for small businesses, lower wages, and even an increased reluctance to hire or promote women of childbearing age.

Two Wrongs Don’t Make a Right

A valid criticism of the Social Security program is that because of the collision between demographic trends and funding methods, the program causes substantial, inequitable income redistribution from the youngest in society to the oldest. In addition, changes in the capital market and standards of living in the United States over the past 50 years mean that the older beneficiaries of the program are overrepresented in the top income quintile. As a result, many see this proposal to provide paid leave benefits through Social Security as a way to reverse some of Social Security’s intergenerational income redistribution, away from retired and older Americans back toward younger ones.

However, the proposal would not actually correct these intergenerational inequities. That can only be done by lessening the existing redistribution from later birth cohorts to earlier ones—either by reducing benefits for older cohorts or by shifting from a pay-as-you-go system to a funded one through, for example, the establishment of personal accounts. The paid leave proposal is instead a proposal for younger Americans to pay for income during their parenting years out of their own eventual retirement income; in other words, if it works as designed, it would have no net effect of correcting intergenerational inequities.

Moreover, owing to Social Security’s currently projected insolvency, changes will need to be made to individuals’ benefit levels and retirement eligibility ages in any event. Would those who have taken parental leave be subjected to these changes on top of the delays they have already accepted in exchange for parental leave income? These are additional considerations the proponents have yet to provide answers for.

Conclusion: The High Cost of Good Intentions

In his book, The High Cost of Good Intentions, Hoover Institution scholar John Cogan explains how throughout history government programs were created for one reason or another, but over time the stated mission was expanded beyond recognition. Federal entitlement programs demonstrate this inevitable tendency. Originally designed to provide a measure of economic security to senior citizens and a safety net for the poor, they now redistribute money to Americans in all income classes, rich and poor alike. As Cogan explains, over 60 percent of all US households that receive entitlement program benefits have incomes above the poverty line before the receipt of those benefits. Cogan further notes that over 30 percent of the benefits go to households in the upper half of the income distribution. The same general pattern is found across federal programs from food stamps to Social Security Disability Insurance to Medicaid.

The same will happen under a paid parental leave policy, especially since it will take decades to implement intended budget offsets. We can, for example, expect future profamily interest groups to argue that beneficiary parents who are later required to delay retirement are being subjected to a “retirement penalty” or a “baby penalty” compared to nonparents. We can also expect other interest groups to demand that the four-quarter work requirement be eliminated so that everyone can benefit, and we expect those interest groups will have statistics at their disposal demonstrating that the work requirements are unfairly regressive. Other possible changes in eligibility might include allowing nonparents and parents alike to tap their future Social Security benefits to pay for college tuition, the care of an aging parent, or the purchase of a house or a car. The Shapiro proposal itself anticipates future expansions, saying that “after the program has existed for several years, policymakers can study its operation and effects and better evaluate whether it would be appropriate to offer more generous benefits or to open up the program to other populations (such as individuals needing paid medical or family leave).” Given that Social Security’s retirement benefits are already inadequately financed, the program simply cannot afford the risk of these myriad benefit obligations being added without reliable assurance that budget neutrality principles will be upheld over all time.

History may not always repeat itself. But it sure does rhyme. If history is our guide, it is unrealistic to expect that any new policy to provide paid parental leave through the Social Security system will remain in its original form and merely redistribute tax dollars to parents in the short run from their future retirement benefits. More likely, it will devolve into another welfare program—funded from a Social Security trust fund that was not designed for that purpose and shouldn’t be used for it now.