Social Security: Investing the surplus in marketable securities

Originally published in TheHill.Com.

It is human nature to look for the easy solution.  My kids do it.  Our politicians do it.
It is one thing when my kids believe the tooth fairy will fix their cavities.  It is another thing when politicians try to convince the public at large that there is no shortage of ways to reduce the growing gap between the projected revenue and expense of Social Security.

There is no free lunch anywhere in the system. If, for example, you remove the maximum taxable wage cap, the system will generate more revenue. At the same time, the change will make it harder to raise revenue for other governmental needs, such as paying down our debt. Increasing payroll taxes is a political statement that Social Security is relatively more important than other governmental priorities.

The Center for Retirement Research at Boston College released a study that suggests it has identified the free lunch solution for Social Security. They believe that we can solve the financial gaps by investing some of the existing surplus in marketable securities.

The research reached an astounding conclusion:

At the 50th percentile of outcomes, equity investing has the potential to maintain a healthy Trust Fund ratio through the 75- year period.

The concept holds the appeal of the easy solution. If we were just smarter about the way we handle the existing resources of the system, all of the difficulties in the program would melt away.
Is it possible? Maybe. The Social Security Administration (“SSA”) doesn’t belief so. Its research suggests that investing 40 percent of the surplus in marketable securities will increase the exhaustion point by a total of 1 year, rather than 75. So success isn’t a sure thing.

The success, or failure, of this concept largely rests upon the investment returns generated by the Trust Fund. Typically, the baseline for most of this type of research is based on historic market returns.

This theoretical assumption isn’t a realistic expectation. Social Security’s revenue is correlated to wages which is correlated to the stock markets. The connection means that Social Security will statistically overbuy market tops and under-invest in market bottoms. It is a statistical anti-dollar cost averaging concept.

Secondly, the index return does not incorporate the cost to buy and sell the security. This problem is largely overlooked today because capital inflows of the typical index fund are sufficient to meet redemptions.

That will not be the case with Social Security. The investment redemptions of Social Security will be priced at the bid.  The size of these redemptions may well move the market down as the government sells its positions. A theoretical return of the index never prices in the spread.

The paper from CRR looks at the returns of the Wilshire 5000, an index that includes highly illiquid stocks. One example is Acme Corporation, ACU, a $20 stock which trades with a spread of $0.50 on an average daily volume of 2000 shares. In other words, the loss starts at 2.5 percent and rises as the price of the stock falls.

This solution does not fix Social Security.  It redistributes the allocation of associated risk. Investing the surplus in marketable security lowers the risk of insolvency by investing in assets with higher returns that introduce credit risk that isn’t present today’s portfolio.

This concept introduces the possibility that Social Security will be adversely affected by a credit market disruption. That risk will be borne progressively by future workers. The possibility that a credit event will occur over the next 10 years is less than over the next 50 years. In this respect, the policy option is no different from the standard line that we will tax future workers more.

Redistributing risk might be a good idea. It isn’t a good idea to base that decision on market returns that Social Security will never achieve.

Social Security deal of 1983 was a punt

The Holy Grail of politics is a ‘grand bargain’ on Social Security in which both parties agree to a mix of benefit cuts and revenue increases that will stabilize the system for the foreseeable future.
Pundits and politicians point to the negotiations which saved Social Security in 1983 as a model of success to imitate.  Obama even invoked the memory of 1983 Reform during the 2012 Presidential Debates.
“Social Security is going to have to be tweaked the way it was by Ronald Reagan and Speaker -- Democratic Speaker Tip O'Neill.”
The story goes that elected leaders put principle over ideology. Bipartisan compromise produced the additional revenue and benefit reductions necessary for the system to not only to survive, but build a $2.7 trillion surplus in the process. It is a wonderful story.
The problem, of course, with this narrative is reality. The 1983 reform of Social Security largely shifted the burden of Social Security from one generation to the next with the consequences falling heavily on late-boomers and those that followed. The Republicans and Democrats came together in a bipartisan effort to kick-the-can.

This legislation wasn’t a hard-choice politically because the consequences were phased-in such that the group most affected were non-voters.  The largest tax increases and the maximum benefit cuts targeted people who were 11 and younger at the time. The political calculus behind the bipartisan compromise is simple: 11 year-olds can’t vote.

The reform contained many changes, but three of these changes accounted for roughly 75 percent of the solution. All of the money-makers disproportionally targeted future generations.
The most favorable change for Social Security was the taxation of benefits. This revision helped the system because the tax revenue collected by the IRS on Social Security benefits was returned to Social Security. The law approximated a means-tested claw-back of benefits by the system from people with “other substantial income”.

This aspect of the reform increasingly affects retirees over successive generations because the trigger thresholds are not indexed to inflation.  The Greenspan Commission originally estimated that these rules would affect “10 percent of OASDI beneficiaries”. 30 years later, the Social Security Administration estimates that nearly 40% of beneficiaries pay taxes on their benefits. Separately, actual IRS data shows that 70 percent of tax returns filed in 2012 that include Social Security benefits generate revenue under this provision.

The next largest change increased gradually the normal retirement age for people born after 1937. The people who felt the full impact of these adjustments were 23 and younger at the time.
The final large impact change was the expansion of Social Security’s coverage to new participants. The law required at all “new” federal employees to participate in Social Security. Again, older workers were largely insulated from the change.

While these changes may seem modest, they weren’t.  Data from the Urban Institute shows that the average retiree couple in 1985 enjoyed $3 of expected benefits for every $1 that they contributed. Today, the average retiree expects to lose money on Social Security. So, benefit reductions were more of an option in 1983 because benefits were relatively generous compared to past contributions.
Separately, Congress was able to create benefit cuts in 1983 that cannot be replicated.  In 1983, the system had no means test. The one that was added in 1983 targeted retirees with an income ($25,000) that approximated the median income of a 4 person family.  Today, these rules hit retirees who earn only slightly more than the poverty-line.

Congress’ options have narrowed considerably as the problems in Social Security have worsened.  As Charles Blauhous points out, “In the early 1980s they merely had to get through a relatively small near-term solvency crisis before entering decades of previously-projected surpluses as the baby boomers moved through the workforce.” (emphasis added)  Social Security does not project any surpluses for an indefinite period of time.

Someone turning 66 today expects live long enough to experience substantial benefit reductions.  It is simply delusional to believe that we can protect people who are 46 and older again.  We have to make changes just so that existing retirees can collect their promised benefits – and that assumes favorable economic conditions. It is no longer economically possible to kick the can.

The story of the 1983 agreement on Social Security is one of Irish drinking buddies who found enough common ground in the good of the country to overcome their ideological differences.  The hard reality is that the president and Congress reached a bipartisan agreement that voters would take more from Social Security than it would give to their children.

Social Security Crash Explained In 4 Simple Steps

When Social Security was created in 1935, the system was designed to be funded by workers not financed by their children. It wasn't a generational-transfer or a Ponzi scheme.  Since that time, the system's finances have deteriorated virtually every year to the point where the financing gap is nearly the size of our entire GDP.

So what the hell happened? How to destroy the future in four easy steps:

1. Give To Voters

The original law included automatic tax increases which would have increased the cost of Social Security to 6% of wages from its 2% base. Over the 1940s, Congress waived every increase, one of which required a Congressional override of FDR's veto. Funding for Social Security did not reach the originally envisioned 6% until the 1960s. Self-employed workers would not pay 6% until the 1970s.
These tax cuts transformed Social Security from a system paid by workers to a system financed by children.

2. Take From Non-Voters 

Once voters did not have to pay for benefits, Congress raised benefits — every election year in the 1950s. Social Security Act Amendments of 1950, 1952, 1954, 1956, 1958 all increased benefits. These increased the value of existing benefits, created new benefits, or expanded coverage to more Americans. The 1950 Amendment raised benefits by 77%, 1952 (12.5%), 1954 (13%), 1956(added disability), 1958 (7%).

A couple retiring in 1960 expected to collect $8 of benefits for every $1 of contribution. Basically, Congress in the 1950s was selling dollars of benefits to voters for little more than a dime. The difference between the cost and the benefit was largely passed on to future generations who had no vote in 1950.

3. Allow The Federal Reserve to Lower Interest Rates At A Cost Of $1.2 trillion of Projected Interest Income (in 2012 alone)

In 2011, projected interest income over the life of the Trust Fund was 3.6 trillion. In 2012, projected interest income over the life of the Trust Fund was 2.4 trillion.

Thanks, Ben Bernanke.

4. Ignore The Problem

Social Security almost reached insolvency in 1983. At that time, Social Security had more than 40 years of promises embedded in a system that did not have a penny to pay them.  The solution to these problems in 1983? Repeat step 1. Repeat step 2. Prepare for step 3. The solution to these problems in 2013? Repeat step 1. Repeat step 2.

Today, Social Security is not much different than spending quarters to buy dimes, provided that  Congress is able to get our kids to spend quarters to buy nickels.