The deep and punishing recession that began with the financial crisis of 2008 will almost certainly become, in tomorrow’s history books, a demarcation line separating what was and what is yet to come.
Since the end of World War Two, the United States and other Western democracies have used their growing wealth and prosperity to build generous social welfare states. These programmatic structures have varied in size and scope, depending on the cirucumstances unique to each country. Still, without exception, the leaders of the rich, industrialized West have sought to improve the security of workers in the postwar years by extending to them better health and retirement income protection, sometimes through direct governmental programs and sometimes through quasi-governmental programs administered by employers.
Although it has been clear for some time now—at least three decades—that these arrangements need to be revised substantially to survive in the twenty-first century, the global financial crash made it abundantly clear that using debt to paper over fundamental imbalances in government finances was not a sensible solution. The social welfare programs that were erected in the postwar era were premised on assumptions of robust fertility rates, perpetually growing workforces, and never-ending economic growth. But without exception, the population of the industrialized world is rapidly aging, birthrates are anemic, workforces are stagnant or declining, and global economic competition has suppressed the wage growth of the West’s middle class.
The United States isn’t exempt from these problems. The Baby-Boom generation is on the verge of retirement, which will swell the ranks of enrollees in entitlement programs. The U.S. workforce is still growing, but not nearly as rapidly as the population age 65 and older. And the middle class has gone through a long period of stagnant wage growth.
The effects of these demographic and economic shifts are already evident in the federal government’s finances. Federal spending has been soaring in recent years and will continue to rise rapidly over the next decade. Meanwhile, the economy is not expected to expand nearly fast enough to produce revenue that keeps pace with obligations. The result is an expected explosion in deficits and debt that will almost certainly precipitate a debt crisis of some sort before too many years pass.
At the center of the looming U.S. fiscal crisis is the nation’s most popular social program, Social Security. Since its enactment in 1935, it has served as the foundation of retirement security for generations of American workers and their dependents. But it is also the federal government’s most costly program, and is set to become much more costly over the coming two decades. Moreover, because the workforce is growing more slowly than is enrollment in the program, revenue has now fallen below annual benefit payments, and under current law, revenue is not expected to ever again exceed annual benefits. The program’s annual cash deficits are projected to deplete the Social Security’s trust-fund reserves over the next two decades. Of course, these reserves aren’t assets in an economic sense. They are simply U.S. treasury bonds and securities issued to the trust fund over the years. Redeeming them will require the government to raise taxes, cut spending elsewhere, or borrow more in order to turn the debt into cash that can be paid to beneficaries.1
One way or another, change is coming to Social Security; it will not be possible to run a cash deficit in perpetuity. The only question is what kind of change will be implemented. Policymakers must properly diagnose what ails Social Security before fixing it. Only then will they be able to put in place reforms that can stand the test of time.
The Unspoken Source of Insolvency
The Social Security program has been the subject of a nearly continual political and policy debate for the better part of fifteen years—although no significant changes to the program have been enacted since 1983. It is now almost forgotten that President Bill Clinton took initial steps toward a Social Security overhaul in the late 1990s, engaging in a series of nationwide “open forums” on the future of the program before abandoning the effort in favor of the more rhetorical—and politically safe—”Save Social Security First!” slogan. While notionally aimed at “saving” the Social Security surpluses, in the end, the Clinton Social Security effort meant little more than “Don’t Cut Taxes!”
In 2005, President George W. Bush, having campaigned on Social Security reform in the 2000 and 2004 presidential elections, attempted to put the issue on the national agenda. His proposal to introduce voluntary personal accounts set off a heated debate among reformers and program advocates, with scores of experts queuing up to advocate a politically diverse range of recommendations, with a number of these recommendations taking form as competing bills before Congress. Despite the intense level of activity, the president was never able to get traction for his ideas, as there was little momentum or consensus for reform.
With so much discussion and political debate in recent years, one might think that every possible diagnosis of and remedy for Social Security’s long-term financial challenges has been offered and debated. Yet there has been very little mention of the central issue in financing Social Security—namely, the long-term fertility rate. Indeed, if the U.S. fertility rate were expected to return to the levels seen in the 1950s and through the mid-1960s, the subject of Social Security reform would likely never come up at all. With higher birthrates, there would be no financing crisis, as the projected workforce in the decades ahead could support the growing numbers of elderly Americans. With no financing shortfall, politicians would gladly leave the program alone.
Unfortunately, fertility is not projected to rise to the levels seen in earlier eras, and, consequently, Social Security does indeed face a substantial long-term financial shortfall. As Social Security again takes center stage in the national debate, policymakers need to take time to understand the critical relationship between fertility and Social Security financing, as well as the potential implications of different reform options for indirectly improving or worsening the American fertility problem over time.
Pension Financing Demographics
Social Security is a conventional “pay-as-you-go” state-run pension system, in which taxes collected from today’s workers are used to pay benefits for today’s retirees. The benefits provided to retirees are based on their earnings records during the course of their working careers. This kind of pay-as-you-go system can be defined with stylized mathematical formula, represented in Chart 1.
The critical assumption in this equation is that, over time, a pay-as-you-go pension system must collect revenue (“financing”) that keeps pace with annual pension-benefit payments. Pension financing can be calculated by multiplying the payroll-tax rate (T), the average earnings of workers on which the tax applies (E), and the total number of workers paying into the system (W). Pension benefits are determined by multiplying the average pension paid (P) by the number of retirees (R).
With these parameters established, it is possible to reconfigure the equation to provide interesting insights into the dynamics of a pay-as-you-go system. The ratio of what someone gets in retirement to what he or she earned while working is called the “replacement rate,” and is depicted by “P/E.” The ratio of retirees to workers is called the “dependency ratio,” and is represented by “R/W.” And the multiplication of these two ratios can be used to calculate the required tax rate necessary to keep the program solvent.
Fertility, of course, is a primary determinant of the number of workers over time. As fertility falls, the old-age dependency ratio increases. As the old-age dependency ratio increases, payroll-tax rates must also rise to maintain pension-financing balance (unless replacement rates are cut). As shown in Chart 1, taking replacement rates of 50 percent and 70 percent as examples, movement of the old-age dependency ratio from .20 to .60 requires a substantial payroll-tax hike—from 10 and 14 percent of payroll to 30 and 42 percent of payroll, respectively. Of course, such high payroll-tax rates would have significant negative consequences for employment and economic growth.
The demographic dynamic of pension financing shown in Chart 1 reflects a prototypical pay-as-you-go state pension scheme. Of course, nearly every country’s state pension has detailed provisions that make it unique and different from the basic structure assumed in the equation. Nonetheless, the main point of the equation presented in Chart 1—that state pension financing is dependent on the demographic forces driving the old-age dependency ratio—is evident in the moves to reform state pension systems around the world to avoid crushing tax hikes.
To be sure, the problems plaguing state pensions are due, in part, to rising longevity. In the United States, for instance, male life expectancy at 65 increased by 5.3 years from 1940 to 2009, and for women it has increased by 6.3 years.2 In much of Europe and Japan, lifespans increased even more dramatically during this period. Unfortunately, the rules of state pension schemes have not changed sufficiently to reflect increases in longevity. The result is that state pensions are paying benefits to workers who are spending an increasing proportion of their lives in retirement.
Of course, living longer is a positive development with a ready remedy for state pension financing—longer working lives. The fall in birthrates is more alarming and harder to address. As shown in Chart 2, the decline in the Total Fertility Rate (TFR)—the average number of live births to a woman in her childbearing years—over the period 1950 to 2000 has been between 0.8 and 2.2 among the world’s richest nations.
The implications of these falling birthrates are startling to consider and beginning to become more evident, particularly in Europe and Japan. As just one of many examples, in Germany, the number of people in their twenties has already dropped by 25 percent since the 1980s. Astoundingly, the population in Germany is expected to drop from 82 million in 2000 to 68 million in 2050.3 The impact of falling fertility on pension financing is clear: fewer workers in the future mean either higher payroll taxes or lower pension benefits. The effect of falling fertility can be seen clearly in the projected increases in the old-age dependency ratios for countries around the world, as shown in Chart 3.
The Inherent Contradictions of Social Insurance
Though policymakers have expended considerable effort in trying to understand and address the difficulty of financing state pensions with rising old-age dependency ratios, they have worried far less about understanding the dramatic fall in fertility. Most policymakers have assumed that fertility is an endogenous factor—a given, determined by factors outside of the state pension structure and government policy. In particular, it is well known that improved medical care for infants and declining infant mortality reduces fertility.
But there is substantial evidence that a particular relationship between public pension schemes and fertility does indeed exist. Foreign as it may sound to the modern ear, a motivation for having children in earlier times was economic security in old age. As parents became frail and less productive, it was expected that one or more of their adult children would take care of them. Married couples thus “invested” in numerous children, in part to ensure the next generation would have the economic capacity to provide for them in their final years. With state-run Social Security schemes, the government has largely absorbed this family responsibility. Married couples have a much diminished economic incentive to have children because now they are counting on—and paying for—government-based old-age support.
Social-expenditure data complied by the United Nations Population Division and the Organization for Economic Co-operation and Development (OECD) reveal that fertility rates in industrial democracies decline as Social Security spending (measured as a percentage of GDP) increases.4 Moreover, a 2005 study by economists Michele Boldrin, Mariacristina de Nardi, and Larry Jones provides strong empirical support for this insight into the fertility dilemma. These scholars looked at long-term trends in Social Security spending and fertility rates around the world and estimated, using regression analysis, that a government-run pension system equal to 10 percent of a country’s economy suppresses the TFR in that country by between 0.7 and 1.6 children, after controlling for other variables. Such a finding is extraordinary given that most industrialized countries now have TFR’s well below 2.0. This research confirms that program size matters. The bigger or more generous the Social Security scheme, the steeper the fertility decline.5
The relationship between social insurance and fertility points toward an internal contradiction in the social insurance model. To finance programs providing for the retired elderly, society needs a growing working-age population, but the presence of the state-based pension benefit—particularly if it is large—reduces the incentive of younger workers to have children. Thus, U.S. Social Security, like government pension plans the world over, is built on a fundamental and poorly understood contradiction: it reduces the economic incentive within a family to invest in children even as it remains ever-dependent on a new generation of productive workers to keep the program afloat.
This insight, it turns out, is not a recent revelation. In a paper presented at the Family Research Council, historian Allan Carlson pointed out that as early as 1940 Gunnar Myrdal, the Swedish economist, had observed in a Harvard lecture that social insurance pensions contained a contradiction.6 For all of human history, adult children were the safety net for the elderly, taking care of aging parents, often in the same homes. With social insurance, the government takes resources from workers to finance direct government assistance in old age, effectively absorbing what was once an entirely family responsibility. While social insurance depends on a productive workforce to pay taxes, families have less of an economic incentive to have children because now they are counting on—and paying for—government-based old-age support.
The fact that Social Security and fertility are inversely related should not be surprising. In fact, the social teaching of the Catholic church—particularly the concept of subsidiarity—warns policymakers to carefully consider the proper balance among societal institutions. According to this teaching, similar to the Protestant understanding of “sphere sovereignty,” higher levels of societal institutions—those farther removed from the daily lives of its citizens, such as a central government authorities—must respect the roles of lower level institutions, not taking for themselves rights and responsibilities that can be properly handled by communities, families, and individuals. This concept of subsidiarity is particularly important for protecting the rights and responsibilities of the family. At the same time, Catholic social teaching allows governments to institute social policies promoting the common good if individual families and citizens would be better off as a group with such a policy in place.
So the question of Social Security is a matter of balance. In modern and wealthy societies, most would agree that it is proper for the government to provide a level of common social assistance and insurance to the elderly to assure that no senior citizen is left destitute in his final years, regardless of his family situation. Yet both society and the economy are stronger when families remain vibrant and independent—producing children and caring for themselves as much as possible.
Although other countries are facing more severe fertility problems than the United States, the U.S. TFR remains well below the levels seen in earlier decades and well below the levels necessary to assure stable financing of Social Security. As shown in Chart 4, in their intermediate set of assumptions, the Trustees for the Social Security program assume that the U.S. TFR will settle at a rate of 2.0 over the long run, which is below the 2.1 TFR needed to replace the population over time.
With relatively low fertility and rising longevity, the gap between Social Security revenue and spending is about to become very wide, and stay that way indefinitely. As shown in Chart 5, Social Security income (payroll taxes plus some taxes collected on benefit payments) will hover just below 5 percent of the GDP over the coming years. Today, Social Security spending is at a comparable level, but as the Baby Boomers retire, spending will soar and never fall back again to its prior level. The 2010 Social Security Trustees’ report projects the excess of Social Security spending over income will reach about 1 percent of GDP and then stay there for most of the coming century. It is inconceivable that the nation could afford to run to such a large permanent deficit in Social Security.
The U.S. Social Security System would be in far better financial condition if U.S. TFR rose from where it is today to something closer to what it was two generations ago, in the immediate postwar era. According to the 2010 Trustees’ report, every 0.1 increase in the ultimate TFR decreases the Social Security deficit by about 6 percent. This means that if the TFR rose from where it is today (just about 2.0) to 2.3 in about 2034, then the deficit in Social Security would fall by 20 percent, thus substantially lessening the need for other benefit reductions.7
Initially, higher fertility reduces tax receipts of the Social Security System, as labor-force participation drop as more women leave the paid labor force to give attention to raising their children. But over the long-term, the workforce expands as a larger younger generation comes of age. As Edward Crenshaw and Kristopher Robison note in their recent study, “If a society can hold its own economically during its baby booms, then it stands a good chance of ratcheting up its economic activity when these children finally enter the labor force.”8 With passing years, the gap between spending and revenue narrows more and more as a larger working-age population matures and boosts the size of the economy.
Pro-Family, Pro-Growth Reforms
The size of the coming Social Security financing gap is so large that it is unlikely that policymakers would—or should—rely entirely on expected fertility improvement to solve the problem. Some programmatic reform will be necessary. The starting point for a sensible approach should be unyielding opposition to any increases in the program’s current size, especially increases financed through tax increases.
The current Social Security payroll-tax rate and wage base—12.4 percent of wages up to $106,800 since 2009—should be ceilings (the taxable wage limit is already indexed to increase with average wage growth each year). Policymakers should resist the temptation to enact a tax hike or to raise the wage-base line to secure a Social Security reform agreement. Such an agreement would only lead to further downward pressure on the U.S. birthrate. Social Security’s financing gap needs to be closed with more robust population growth and benefit adjustments.
The best way to alter Social Security benefits is to redesign how the system operates so that it automatically adjusts with evolving demographic reality. When Social Security began, male retirees at the age of 65 could expect to receive benefits for an average of about twelve years. Today, men at the age of 65 will receive benefits for an average of sixteen years. And by 2050, the average male retiree is expected to live about 19 years after reaching age 65. A similar trend is underway for women.
Private insurers selling annuities carefully calibrate the monthly payments they are willing to provide for a fixed premium based on up-to-date mortality data to keep annuity costs in line with the purchase price. Social Security, however, will finance longer and longer retirement periods for each new generation of retirees—even though the different cohorts will all pay the same payroll-tax rate during their working years.
One way to deal with this problem would be to set the Social Security normal retirement age administratively, as determined by up-to-date life-span data. Congress could give administrative authority to the Social Security Administration to set the age after establishing a uniform number of retirement years, with actuarially fair reductions for early and delayed retirement. Once enacted, this approach would remove “raising the retirement age” from on-going political consideration, so shielding political leaders from unwelcome pressure on the issue.9
The normal retirement age is already scheduled to increase from age 65 to age 67 by 2027, but it needs to go to age 68 to keep the number of years in retirement consistent with current levels and keep pace with expected declines in mortality. To improve Social Security solvency, a more aggressive reform may be necessary. The Congressional Budget Office estimates that moving the normal retirement age up to age 70 over the next fifty years—thus reducing the targeted time in full retirement to the levels experienced by persons retiring in the 1970s—would cut Social Security spending, by about 12 percent when fully implemented.
Perhaps more important, Social Security benefits should be adjusted based on shifts in the old-age dependency ratio. As the number of retirees to workers increases, benefits will need to be adjusted downward to prevent the need for tax increases. But, conversely, if birthrates improve over time, building an automatic adjustment into the benefit formula can mitigate benefit restraint automatically as the dependency ratio improves.
The United States could model this particular reform along the lines of what Germany has implemented in its retirement system.10 Germany’s “sustainability factor” measures the change in the ratio of beneficiaries to workers contributing payroll taxes on a yearly basis. As this ratio increases in the years ahead, a weighted-adjustment factor could be applied in the U.S. system to the initial Social Security benefit formula, slightly reducing the benefits payable to a defined group of new retirees. For example, today the old-age dependency ratio is .22; it is expected to reach .36 in 2050. A new dependency-ratio adjustment factor would convert this shift into a uniform percentage reduction in the initial, full-benefit retirement formula. To minimize annual fluctuations, the factor could be calibrated based on a ten-year moving average. Once retirees start receiving benefits, they would no longer be subject to this adjustment factor and would be protected from inflation through annual cost-of-living increases.
Once they have implemented these automatic adjustment mechanisms, policymakers should reverse the negative impact of Social Security on fertility rates. Today, two workers with identical wage histories pay the same contributions and get the same benefits, even if one of them has numerous children and the other has none. Any discussion of fixing Social Security must focus on correcting this imbalance, making the reform plan more appealing to young families.
The easiest way to fertility-enhancing reform would be to build directly into the Social Security formula a factor reflecting the value to Social Security of raising an additional future contributor. Like the adjustment in the retirement age, this adjustment for having and raising children could be calculated by the Social Security Administration and adjusted over time as circumstances warranted. The calculation would reflect the additional value to the Social Security of families raising children compared to those that do not. The calculation would reflect average lifetime earnings and benefit payments for families over time, giving appropriate additional benefits to workers who have reared children—the future contributors that the Social Security System must have to survive. Because parents with multiple children would be raising multiple future taxpayers, the calculation would show the expectation of a net gain to Social Security from such a family compared to one without children. That net gain could then be partially shared with the parents during their retirement years in the form of higher monthly benefits.
There are, of course, many non-economic reasons for declining fertility in the United States. Societal norms, the decline in religious practice, and later median age at first marriage all play roles. Social Security reform by itself will not be able to reverse those trends. But there is little doubt that Social Security is partially responsible for the demographic decline that has been underway since the 1970s. Whatever can be done through changes in Social Security to boost fertility should, therefore, be a top priority of reform. Such changes would not only help restore solvency to Social Security but also help ensure that America in the twenty-first century remains as vital and dynamic as she was in the last.
Mr. Capretta, a fellow at the Ethics & Public Policy Center, was an associate director of the U.S. Office of Management and Budget, 2001 to 2004.
6. Allan C. Carlson, Making Social Security Reform Family-Friendly,” The Family in America, April 2005, <http://www.profam.org/pub/fia/fia_1904.htm>.
8. Edward Crenshaw and Kristopher Robison, “Socio-demographic Determinants of Economic Growth: Age-Structure, Preindustrial Heritage, and Sociolinguistic Integration,” Social Forces 88.5 (July 2010): 2217-40.