On Monday, the trustees of the Social Security system released their annual report documenting the system’s current and projected health. The key message of the report was that the Social Security trust fund is in deteriorating condition and the Old-Age and Survivors Insurance (OASI) program – the portion that covers some 94% of American workers – will exhaust its trust fund by 2033. Various media reports presented this as a surprise – especially since the report from last year had a D-day of 2036. How could we have lost three years off the trust fund in a single year?
For anyone who has followed the health of the system and knows its history – or who can spot the parallels between the U.S. and what is happening in Europe or more quietly in Japan – this is no real shock. The table below shows the projected exhaustion date for the OASI fund as well as the cumulative shortfall of forecast OASDI payroll taxes (Old Age, Survivors + Disability Insurance) as a percentage of total payrolls for the past six Trustees’ reports
Over the past five years the exhaustion date of the OASI trust has been brought forward by eight years. The payroll tax shortfall percentage, or essentially how much larger the payroll tax would need to be to bring the program into balance, has increased from 1.7% to 2.7%, the largest actuarial deficit since before the 1983 Social Security amendments that raised normal retirement ages from 65 to 67 for younger cohorts. We are heading in the wrong direction.
What is causing the shortfall? Setting aside the actuarial p’s and q’s, the core of the problem comes down to mortality, demographics and growth.
Mortality improvements are well-known. At the time of the Social Security system’s creation in 1935, life expectancy from birth was roughly 62 years for males and 66 years for females, according to the Congressional Research Service, whereas those figures are currently closer to 75 and 80. This well-known phenomenon of mortality improvement through time is incorporated into the forecasts, but to the extent that survival rates are understated, this could be a potential source of risk. Much of the reason for the changes to the benefit structure in 1983 can be tied to an under-allowance for increased longevity. Indeed, in the nearly three decades since these changes, life expectancy has increased by a little over three years for men and approximately two years for women, explaining part of the funding slide.
Demographics matter a great deal for a pay-as-you-go system. As far back as 1958, the late Nobel laureate Paul Samuelson recognized that fertility and immigration were key to a vibrant and healthy Social Security system. Growth in the labor supply and, consequently, payroll taxes effectively acts like an increase in the discount rate of future liabilities. Extrapolating then-current birth rates in the midst of the Baby Boom, the prospects for Social Security looked quite rosy. However, what Samuelson failed to see then, and what confronts us now, is that the baby-producing boom subsided, and we were left with a cohort moving through the system that was proportionately larger than following tax-paying generations. Again this has been known for decades.
What is less well understood is the interplay between the trend seen in the wake of the Baby Boom and growth. This is the risk we face going forward. A growing academic literature has tied growth – economic and financial – to demographic trends, and the implied forecasts do not look pretty. As seen in more mature countries, when the ratio of retirees to workers rises materially, growth slows, financial returns disappoint, and real interest rates decline and stay persistently low. This raises some potential questions about two key growth rates in the Social Security Actuarial model: productivity growth and real interest rates.
In their intermediate, or baseline, forecast, the actuaries assume a 1.68% long-term productivity growth. That growth rate is roughly equal to the 1.8% growth in long-term productivity seen from 2007 to 2011, according to Bureau of Labor Statistics data on the non-farm business sector, but below the 2.1% to 2.5% rate seen from 1990 to 2007 when the Baby Boomers were in the peak productive years. It’s still higher than the 1.1% to 1.4% seen from 1973 to 1990. If the demographic-driven growth models are right, actual results may disappoint.
The real interest rate assumption may miss the mark by a wider measure. The Trustees’ report baseline assumes a 10-year forward real interest rate of 2.9% – the same rate used in 2006 when yields on Treasury Inflation-Protected Securities, or TIPS, were more than 2% higher. Current market-implied forward long TIPS rates (PIMCO calculations on Bloomberg data) are closer to 1.4%. If discounted by the lower market-implied rate, the present value of future benefits would be materially higher and the projected shortfalls significantly greater. One insidious consequence of so-called financial repression (i.e., keeping rates so low for so long that you create negative real returns on savings and destroy wealth) is that while it reduces the real future servicing cost of the government’s marketable securities, it raises the present value of future social insurance obligations. As these contracts are currently inflation-linked, barring restructuring it is not clear which of these effects will dominate.
If the trend in the nearby table continues – and given our long-stated views for slower growth in the developed world, we would contend that it is more likely than not that it will – it is quite possible that the OASI trust fund could run out of money in little more than a decade rather than the current projection of a little more than two decades.
So what can be done?
The answers have changed little from the policy choices studied by the Greenspan Commission in 1983, Bush’s Commission to Strengthen Social Security in 2002 or Simpson-Bowles in 2010. The answers lie in some combination of reduced benefits in the future, higher retirement ages or higher taxes. Cutting benefits by the 25% (or more) required to bring the system into balance is neither politically viable nor economically rational. The most likely result will be a little of all three – lower indexation of wage growth for benefit determination, delayed retirement for those set to retire in 10 or more years and higher taxes for everyone. The powers that be in Washington need to wake up to reality. The deficit is too large to be balanced on the backs of the One Percent. Cutting taxes and praying for higher growth may make for appealing political rhetoric but would likely fare no better. To quote the conclusion of the Trustees’ report:
Lawmakers should address the financial challenges facing Social Security and Medicare as soon as possible. Taking action sooner rather than later will leave more options and more time available to phase in changes so that the public has adequate time to prepare.
Amen. The longer we delay, the worse the problem will be. Whoever wins the White House in November needs to make fixing Social Security and Medicare, which is far more complicated and daunting, top priorities.