The Magazine

Gunnar Myrdal Was Right

Social Security's fertility problem.

May 7, 2007, Vol. 12, No. 32 • By JAMES C. CAPRETTA
Widget tooltip
Single Page Print Larger Text Smaller Text Alerts

Barring a political earthquake, President Bush will leave office without achieving his goal of transforming Social Security. That's too bad. A successful Social Security effort would be a significant down payment on much needed entitlement reform. But sooner or later, Social Security will find its way back onto the public agenda. The program's financial problems are simply too big to set aside indefinitely.

In a nutshell, Social Security's long-term prospects are bleak because of rapid and unprecedented population aging. Part of the demographic story is, of course, good news about longer lives. When Social Security started, the average man could expect to live about 12 years after reaching retirement at age 65. Today, he can expect to live 16 years.

But by far the most important factor in aging over the long run is falling fertility. Societies that do not produce children will, quite naturally, grow older. And there is simply no bigger problem than a low birth rate for a conventional "pay-as-you-go" pension system like Social Security. Without a steady stream of "payers," the system cannot "go."

What is not widely understood--despite the work of analysts like Allan Carlson, Phillip Longman, and John Mueller, among others--is that Social Security itself contributes significantly to the problem of low fertility. Odd as it may sound to the modern ear, a primary 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, oftentimes by bringing them into their homes. Married couples thus "invested" in numerous children, in part, to ensure there would be family members to care for them in their twilight years. With state-run Social Security, the government has largely assumed this family responsibility. Married couples have a greatly diminished economic incentive to have children, because now they are counting on--and paying for--government-based old age support.

This insight is neither new nor conceived of by conservative opponents of the welfare state. As noted frequently by Carlson, Gunnar Myrdal, the eminent Swedish socialist economist, observed in the 1940s that state-run, pay-as-you-go pension systems are built on a fundamental "contradiction": They reduce the economic incentive within a family to have children, even as they remain ever dependent on a new generation of productive workers.

The evidence for Social Security's role in fertility decline is compelling and has been confirmed by numerous researchers. A 2005 study by economists Michele Boldrin, Mariacristina De Nardi, and Larry Jones found that a government-run pension system equal to 10 percent of a country's economy correlates with a reduction in the Total Fertility Rate (TFR)--which measures the average number of births per woman during her lifetime--of between 0.7 and 1.6 children, after controlling for other variables (no one suggests government pensions are the only reason fertility has dropped).This is extraordinary given that most industrialized countries now have TFRs well below 2.0. Importantly, this research indicates that program size matters. The bigger the Social Security scheme, the steeper the fertility decline.

Japan, now the world's oldest society, presents a striking case study. Government spending on Japan's state-run pension system increased dramatically in recent decades, from about 1.3 percent of national income in 1965 to more than 12 percent today. During this same period, Japan's TFR fell from about 2.0 to 1.3. The economic consequences of this plunging birth rate will be severe in the years ahead, as the workforce contracts significantly. Japan's working age population--those ages 20 to 64--is expected to decline from 79 million in 2000 to just under 50 million in 2050, a 37 percent drop. With a smaller workforce expected, Japan has been forced to reform its Social Security program three times since 1994 and has scheduled a 0.354 percentage point payroll tax rate hike every year between 2004 and 2017--increasing the tax from 13.6 percent of wages to 18.3 percent. The evidence suggests that this tax increase will only serve to further suppress the one variable that could eventually bring Japan back from the brink--more children and, therefore, more future workers.

The pattern is similar, if not as dramatic, throughout the developed world, including in the United States. The Total Fertility Rate fell precipitously as Social Security expanded in size, from a high of 3.5 in 1955 to a low of 1.8 in 1975. It has since recovered slightly to 2.0. Today, France has a TFR of 1.9, the U.K. is at 1.6, and Germany, Italy, and Canada all have TFRs below 1.4.

The crucial role that fertility plays in pay-as-you-go systems is evident in the future projections for Social Security. Current mid-range estimates assume the U.S. TFR will remain at 2.0 indefinitely, but the resulting modest growth in the U.S. workforce is insufficient to keep up with the ever-growing number of retirees. By 2080, the ratio of the working age population to the elderly will be just 2.4 to 1, down from 4.9 today. A gap between Social Security spending and tax collections is expected to open up beginning in 2017 and widen substantially in the ensuing years, reaching about 1.8 percent of GDP in 2080.

But the future of the U.S. fertility rate is not set in stone, and further improvement would greatly ease the financial strain. A steady rise in the TFR to 2.8--last seen around 1965--would eliminate about one-half of the projected financial shortfall in the Social Security program over a 75-year period. On the other hand, a fall in the U.S. TFR to the levels found in most of Europe would open a much wider gap in Social Security's finances and precipitate an even deeper crisis in the program.

Acknowledgment of Social Security's role in fertility decline is not an argument for abandoning government-sponsored old age support. The elderly--and their adult children--far prefer financial independence to dependence, and Social Security provides important protection against destitution, a common problem in earlier eras for the elderly with no family support. But reformers must understand the population contradiction at the heart of Social Security if they are to think clearly about the future of the system.

The starting point for a sensible reform of Social Security should be opposition to any proposal that would increase the program's current size--including proposals to initiate personal savings accounts with "add-on" contributions. The current Social Security payroll tax rate and wage base--12.4 percent of wages up to $97,500 in 2007--should be ceilings (the taxable wage limit is already indexed to increase with average wage growth each year). The temptation among some Republicans to accept a tax hike to get a bipartisan Social Security deal should be resisted; it would only lead to further pressure on the birth rate. Social Security's financing gap will need to be closed with benefit adjustments, such as increases in the retirement age and reductions in the benefit formula for higher wage earners.

But reform should not stop there. Today, two workers with identical wage histories pay the same contributions and get the same retirement benefits, even if one of them has numerous children--with all of the expense that entails--and the other has none. That's not fair or prudent, as it undervalues the program's need for investment in human capital. Longman, Carlson, and others have suggested reforms that would begin to correct this flawed accounting. These proposals would give workers who are raising a family either a payroll or income tax break or a boost in their retirement benefit to compensate for the costs of investing in future taxpayers.

The point for conservatives is that any discussion of fixing Social Security should properly focus on strengthening young families, which will be both more popular with voters and better policy than approaches that ignore the fertility problem. Placing the cost of any reform within Social Security--a child credit that lowers the payroll tax, for instance--will also ensure there is no increase in the program's unfunded liabilities, as any added costs would have to be offset with benefit reductions, which should also boost fertility. Incorporating automatic, ongoing adjustments in the benefit formula over time, based on actual changes in fertility (as well as longevity), could help sell the reform in Congress. Proponents could then argue that benefits would automatically increase if the fertility rate improves as hoped.

No one should be under any illusion, however. Reversing the long-term slide toward smaller families will take more than a creative Social Security plan. My Ethics and Public Policy Center colleague John Mueller has noted that higher fertility accompanies more frequent religious practice, for instance. Clearly there are powerful social trends at work that can only be addressed with cultural renewal. But Social Security is contributing to the problem, and its reform should be part of the solution.

James C. Capretta, a fellow at the Ethics and Public Policy Center, was an associate director at the Office of Management and Budget from 2001 to 2004.