Recession could cut life of Social Security trust fund by five years

I was asked yesterday how the current recession, which has significantly reduced the Social Security surplus, might affect the life of the trust fund. I put together some back-of-the-envelope numbers that at least guestimate the effects. These are necessarily approximations and mix CBO and SSA projections together (CBO, because they have they latest data available; SSA, because the Trustees projections for solvency remain the most prominent and well-understood). Take them with a grain of salt, but qualitatively I believe they should get in the ballpark.

Here's how CBO summarizes the differences in Social Security cash flow – meaning tax income (payroll tax revenues plus income taxes levied on retirement benefits) minus benefits and administrative costs – between their March 2008 budget projections, which also formed the basis for their August 2008 long-term Social Security projections, and their current March 2009 projections:

Even as of 2018, cash flow is almost $50 billion per year lower than was previously projected. I'm assuming this difference declines to zero within 10 years following 2018, although it's possible that the effect is permanent.

Next I calculate the present value of these new-found losses to tax income. Discounted at a nominal interest rate of 5.4 percent, which is a rough average of the 10-year Treasury bond rate from CBO's latest economic projections, this produces a present value as of 2009 of $621 billion. In other words, these future declines in payroll tax revenues are roughly equivalent to the trust fund balance simply being $621 billion lower as of today.

Next, I compound this present value forward to the year 2040 at a 5.2% interest rate, which is closer to the longer-term rates used in CBO's August 2008 Social Security update. This produces a value as of 2040 of $2.99 trillion. This shows the rough lost value to the trust at the point in which the fund is near exhaustion. In CBO's projections, the program is running annual nominal deficits of around $550 billion at that point (that's around $300 billion in today's dollars, for anyone interested).

Now, I simply divide $2.99 trillion by $550 billion, which gives me a value of 5.45. This indicates that the lost income to the trust fund today is worth around five and one half years of solvency in the 2040s. The Social Security Trustees currently project the program will become insolvent in the year 2041, so the current recession could push that insolvency date forward to around 2036.

Bear in mind that these numbers are approximate – you wouldn't want to do brain surgery this way. We will know more when the next Trustees Report is released. Moreover, lower wages today due to the recession can lead to lower promised benefits in the future, although I suspect the offset won't be on anything close to a one-to-one basis. (Losing a year of earnings due to unemployment would reduce benefits only if that year were among the worker's highest 35 and if he doesn't make it up by extending his work life later. In addition, if unemployment is concentrated among low earners, their benefits may become more progressive and partially make up for any losses.)

In any event, though, this gives a rough feeling for the scale of the effects of the recession on long-term Social Security financing.